Golden Prospects in Papua New Guinea: K92 Mining's Bold Expansion

Executive Summary

K92 Mining Incorporated (“K92”) (TSXV:KNT, OTCQX:KNTNF) is a Canadian company that has been advancing the Kainantu project in Papua New Guinea since 2015 after it had purchased it from Barrick Gold Corporation (“Barrick”). The market capitalisation is C$1.4 billion (US$1.08 billion), and on a fully diluted and adjusted basis the Enterprise Value (“EV”) is C$1.53 billion (US$1.15 billion).

Kainantu is now one of the lowest cost gold mines globally, with production planned to increase from levels of 82,000 ounces per annum (“ozpa”) in 2019 to target 140,000 ozpa in 2021 as the processing plant capacity doubles. K92 is also installing a twin decline that will enable production tonnages to increase from current levels of 200,000 tonnes per annum (“tpa”) to 1 million tpa by 2025, which would take gold production to 318,000 ozpa. The twin decline is being sized with a long-term vision of producing 3 million tpa in the future – a further expansion option which is not considered in this report.

In parallel with the decline development K92 is targeting resource growth at a variety of different scales. On the known veins, infill and expansion exploration is ongoing, to support mine plans and resource estimates. Within the mining lease the Company estimates that the established veins only represent one fifth of the veins that are present, so exploration is ongoing to target new resource areas. Finally, the mining lease sits within a broader 700 km2 project area, and regional exploration is ongoing targeting other vein systems and porphyry deposits.

The goal of the company is to grow earnings by increasing low cost production, and to grow capital value by continuing to add economic ‘ounces in the ground’ through exploration. The management team has done an excellent job of execution on the plan thus far, having more or less created a billion dollar company from scratch in five years. It should be noted, however, that the company is a one-trick pony for good or ill. Even the Company name reiterates the focus… Kainantu, K-nine-two, K92 (geddit? Ed.) The low cost of resource ounce discovery, the geological endowment of the mining lease, the high grade asset all mean that any other project will struggle to match the return of investing in Kainantu itself. Diversification, by virtue of Kainantu’s uniqueness, is largely off the table, which in some ways is no bad thing. The management team is liberated from all of the distractions of trying to build a diversified business, and it can channel Steve Jobs and ‘do one thing well’. The skill set required of a management team tasked with development and exploration of a single project site is specific, and the current management team look up to the task. So far so good.

The flip side of focus is a lack of diversification. A single asset company in a frontier market will trade at a discount to full value due to the risk associated with it being a single asset. The revenue stream from a single asset is binary in that it is either flowing, or it is not. Strikes, pestilence, plague, earthquakes, revolution can all happen (Covid-19 anyone?), let alone commonplace ground failure, community protests and blockades or labour dispute. For a diversified company the impact of one mine closure can be mitigated by support from other divisions. Not so for K92, with all of its eggs in the Kainantu basket, with everything hinging on the safe expansion.

As Kainantu matures as a project the pressure to mitigate risk via a merger will grow, and Crux Investor feels that a good outcome for shareholders would be for K92 to end up as part of a bigger company. But there is a lot of runway at Kainantu before then.

The EV per ‘mineral inventory’ ounce figure is US$353/oz, which is probably about right for a single-asset producer. The high EV/oz figure reflects both the high grade and high margin production, and also the prospect for significant resource growth. Crux Investor believes that the

quality of the asset, the mineral endowment and the potential for spectacular discovery justify this valuation. Furthermore, when Crux Investor runs the slide rule over the Kainantu plans, making what we believe to be prudent adjustments, the NPV emerges at US$1.75 billion, which is significantly higher than the current EV of US$1.15 billion.

Even accounting for a single-asset discount, the valuation of K92 at an EV/NPV of 0.66x is, in our opinion, harsh. Crux Investor feels that with continued quarterly performance the discount will be unlocked, and a reasonable level would be 1x of NPV, which would suggest a 52% rise in the current share price. Note that the Tier 1 gold producers are currently trading at around 1.3x NAV, and that the NPV figures calculated by Crux Investor do not include the option value of future resource discoveries.

To arrive at the NPV figures, Crux Investor undertook a desktop valuation, reviewing historic data, and using industry norms to critique expansion plans. There are significant technical challenges ahead, and when working through the data Crux Investor found there were a number of Red Flags in the plans and assumptions that will hopefully be managed and resolved. Equally the asset is high grade, and producing low cost gold at a high margin, so there is much to like, and many Green Lights. The building blocks of the report are shown below.

When K92 Mining bought the asset, the project area included a defunct mine complete with all infrastructure, including a processing plant. The mine was closed down not because it had depleted all its mineral resources, but because of operational difficulties experienced by Barrick, and probably also because it did not have the scale of production to be of interest to Barrick. Management reports that Barrick viewed the exploration potential of the larger licences as the real prize. Initially the acquisition consideration was set at US$2 million, followed by up to US$62 million over 10 years based on additional discovered reserves.

After refurbishment of the processing plant, K92 started production on a very small scale in 2016 from Irumafimpa, the deposit previously mined by Barrick. Irumafimpa was high-grade, but it presented mining problems including being a very narrow structure with poor ground conditions. Fortunately, exploration drilling had discovered wider, higher grade veins adjacent to Irumafimpa, referred to as Kora, which was almost immediately identified as the area on which to focus.

K92 started in 2016 developing underground towards Kora, and the development also provided drill platforms to enable exploration of the previously untested area between Kora and Irumafimpa. The company planned for production from Kora by mid-2018, which target it easily beat as commercial production was declared in February 2018.

Further good news came in 2019 when it was announced the contingent payment of up to US$60 million was replaced by a fixed amount of US$12.5 million. This was paid to Barrick on 23 August 2019.

The production performance has seen a dramatic improvement in grade with mining at Kora, initially at 10 g/t gold equivalent (“Au Eq.”), improving to almost 21 g/t Au Eq. in 2019 settling at 15.6 g/t Au Eq. for the first half of 2020. Copper and silver are present in the veins, and they are valuable by-products contributing to the economics of the operation. At such grades the operation is a real money spinner. Unlike so many of the gold mining operations in the industry that are of dubious long-term economic value, Kainantu has become, a “veritable gold mine” in both the literal and the metaphorical sense.

The latest resource estimation arrives at mineral resources of more than 15 million tonnes at a grade of almost 10 g/t Au Eq, using a very low cut-off grade of 1.0 g/t Au Eq. However, Crux Investor ‘Red Flags’ this resource estimate for a number of reasons. The concern here is that a large majority of the resources are in the Inferred category and the grade estimation relies heavily on underground sample grades in an area that seems to have a more consistent grade than away from the sampled area. Furthermore, the consultants did not top-cut very high composite grades, which could result in serious overestimation of gold content. The PEA itself notes, for example, that a top-cut applied to just four samples in one particular vein would account for a drop in resources of 0.7 million ounces (“Moz”). And yet no top-cut is applied to the resource estimate. There is risk here, but the operation has a history of positive resource reconciliation performance which is a testament to the quality of the asset. [More gold is produced than the models predict.]

In the PEA, the consultants present a mine plan which would allow for mining of 10 million tonnes, assuming a cut-off grade of 5.5 g/t Au Eq. This “mineable inventory” processed at the current rate of 0.2 million tonnes per annum (“Mtpa”) would yield a life of mine (“LOM”) of 50 years. To sweat the Kainantu asset faster management now aims to increase the processing rate to 1.0 Mtpa, which is being dubbed Phase 3 Expansion. Net free cash flow will benefit from cost reductions through economies of scale.

On paper the expansion makes eminent sense. The challenge is to achieve this from narrow deposits with difficult ground conditions. The main risk associated with K92 is that actual mining conditions may dictate a change in plan and a scaling down of the company’s ambitions. Confirmation of the mine plan is now the subject of numerous studies the most important of which is geotechnical work. The company is currently working to increase the number of sub-levels in the mine to eight by year end 2020. By having more operating sub-levels, the Company will have more operational flexibility and more throughput (by virtue of having more faces to mine and bigger stopes) to increase the amount of ore mined.

This valuation has given the production ramp up the benefit of the doubt, accepting the doubling of production by 2021 (i.e. treating 0.4 Mtpa), ramping up further after 2022 to reach a rate of almost 1.0 Mtpa in 2025. What Crux Investor does not accept is the suggested operating cost rate of US$87/t when reaching the target throughput rate as it is too far below the actual US$304/t in H1 2020. With much uncertainty about what underground mining method will prove practical and the degree by which costs are fixed, this valuation has made a guestimate of the unit production cost of US$170/t which is 70% higher than proposed by the preliminary economic assessment (“PEA”) study, but also 44% lower than current unit cost.

One Red Flag noted by Crux Investor was that the PEA did not calculate the economic performance on an after-tax basis. The omission is a glaring shortcoming as this item will be the major cash drain after (Crux Investor-upward-adjusted) operating costs. Taxes and death are the only certainties in life; why ignore them? Not including the post-tax figures only diminishes the reputation of the consultant, and by implication the Company for not having insisted on the real-world conclusion. Certainly, the project is robust enough to carry the burden.

At the gold spot price of US$1,899/oz on 2 October 2020 the project has a cash operating margin of 74% which is extraordinarily high. The spot prices compared to the gold price of US$1,500/ oz used in the PEA, more than makes up for the upward revision of operating cost and the introduction of taxation, which includes 30% income tax and 15% on expatriation of funds, and accounting for corporate overheads. The pre-tax net free cash flow attributable shareholders of US$2.84 billion is almost exactly the same as the pre-tax US$2.86 billion cash flow shown in the PEA. Discounted at 7.5% to reflect the “new project” status of the production expansion gives an after-tax value of US$1.75 billion. With the extremely high cash margins the sensitivity to changes in gold price, operating cost and capital expenditure is very low. Even at a discount rate of 10%, the NPV gives a value of US$1.51 billion.

To arrive at a full NAV for the Company, Crux Investor should add the option value of the broader mining lease and exploration licences to the NPV figure for the expanded mine. We acknowledge that there is the considerable blue-sky potential from additional discoveries in the Kainantu project area. Given the obvious considerably size of the mineralised system in the Kainantu project area, evident from the numerous targets identified there, Crux Investor is of the opinion that it is highly likely the company will find additional resources in due course. In a sense, though, that is ‘in for free’, and the company can fund it all from cash flow.

As noted above, the company’s fully diluted Enterprise Value of K92 at the prevailing share price on 2 October of C$6.88 is, at US$1.10 billion, clearly lower than the calculated intrinsic value. Crux Investor believes that the haircut is too severe, and we expect the discount to be unwound as K92 Mining continues to create money by mining gold, and discovering more economic ounces that can be mined at a high margin in the future.

Introduction

K92 Mining Incorporated (“K92”) (TSXV:KNT, OTCQX:KNTNF) is a Canadian company which was incorporated in Canada in March 2010 under the name Ottenburn Resources Corporation with the expressed strategy “to explore mineral properties to a stage where they can be developed profitably or sold to a third party”. It first acquired a Canadian prospect Adams Plateau in British Columbia, Canada, but exploration results must have been disappointing as no work ceased after August 2013.

In 2014 and 2015 the company restructured acquiring a British Virgin Islands registered company, K92 Holdings International Limited (“K92 Holdings”), conditional upon a number of preconditions, among others, the effective acquisition of the Kainantu project in Papua New Guinea from Barrick Gold Corporation (“Barrick”). The consideration payable was US$62 million, of which US$2 million initially and the balance over 10 years and based upon additional reserves being discovered at Kainantu. However, during 2019 the purchase agreement with Barrick was amended whereby the contingent payments were revised to a fixed payment of US$12.5 million, which was paid to Barrick on 23 August 2019.

With the acquisition of the Kainantu project K92 inherited an underground mine that operated from 2004 to 2008, mining the Irumafimpa gold deposit. The mine was closed down not because it had depleted all its mineral resources, but because of operational difficulties experienced by Barrick, and probably also because it did not have the scale of production to be of interest to Barrick. In fact, the mineral resources at Irumafimpa and the adjacent Kora area were estimated at the time to amount to approximately 2 million ounces (“Moz”) gold equivalent at a grade of approximately 12 g/t Au Eq.

The property was well supported by regional infrastructure and contained all the necessary site infrastructure for mining operations, including a plant located 7 km from the mine portal servicing Irumafimpa. Refurbishment of the processing plant was completed in September 2016 and the first batch of underground ore from Irumafimpa was treated in October 2016.

Whereas the in-situ grade at Irumafimpa was very high, the feed grade reported from mining such areas was a fraction of this grade. For example, a comparison in 2008 between defined reserves in certain stopes averaging 20 g/t Au was very different from the 8.9 g/t as per grade control estimates. Worse combined with a drop of more than 40% in expected plant feed, the contained gold in the feed was estimated to be 30% of the “reserves”. In addition, the grade control grade compared very high to the back calculated mill feed grade of 5 g/t Au. No wonder Barrick wanted to throw in the towel!

Grade Reconciliation Explainer

Grade reconciliation is the science surrounding what miners think should be in any given tonne of ore, what is actually recovered from that tonne of ore, and reconciling, understanding, rationalising, and minimising the difference between those two figures over time. In essence the aim of ore grade reconciliation is to identify, analyse and manage variance between planned and actual results in a way that highlights engineering or operating opportunities. Naturally this practice covers everything from the initial resource estimation process through to process efficiencies. It is important to get it consistent, and to improve operating efficiency with experience.

There is, of course, complexity. A mine is developed in many stages, and to help understand the process in its entirety, operating teams, when trying to get a holistic understanding of grade reconciliation, divide the process into three broad categories: temporal, spatial and physical.

Temporal reconciliation compares performance across the mining operation on time intervals such as shifts, days, weeks, months, years etc. Spatial reconciliation measures the absolute performance between predictive models and the actual results determined by mapping and survey measurement. Physical reconciliation is concerned with attributes such as contained metal, various quality parameters and volumes.

Ore Grade Reconciliation is checked at various data points along the journey from Ore block estimate to finished product, with Figure 1 showing the reality check stages between the various layers, or each stage of the process from Estimate to Production. Figure 1_1 shows how these classifications of reconciliation relate to the various activities of the mining value chain. The pyramid is inverted because the quantum of information available to mining personnel grows as one progresses up the pyramid.

A good reconciliation process is expected to identify discrepancies as they occur, and proceed to identify and correct the underlying causes. It should be able to lead to the economic quantification of the identified misbalance / discrepancies (how much money is being lost), and therefore to highlight a better way of doing things.

Value is created if the mining company implements methods to create better estimates, improved designs, tighter and more accurate plans and schedules, improved mining techniques to minimise ore loss and dilution, and if it identifies ways to increase metal recoveries during the extraction processes. The ability to measure and analyse data efficiently and consistently enables an operator to design and implement process improvements across the entire mining value chain, reconcile differences between ore block estimates and final production. Though challenges encountered cannot be completely eliminated, they can be adequately mitigated.

Mining is not an easy game!

It was probably for the above reasons K92 focused on developing the adjacent Kora area with much wider veins for production concurrent with intensive exploration drilling of the Irumafimpa and Kora deposits. In February 2018 K92 declared commercial production at the Kainantu mine and mine production focused on the Kora North area. Upgrades of site infrastructure continued in 2019 and 2020.

Historical Operation & Financial Performance

Table 2_1 gives the historical operational and financial performance from 31 July 2015, the year when K92 became the operator of the Kainantu project, until 30 June 2020.

Table 2_1 shows that operationally:

• Plant production consistently increasing reaching almost 0.2 Mtpa in H1 2020 on an annualised basis.
• The feed grade has, over time, fluctuated between 10 g/t and 21 g/t gold.
• Metallurgical recovery is in the lower nineties for both gold and copper.
• By-product revenue is minor at around 3%.
• Site operating cost is approximately US$300/t milled.

Table 2_1 shows that K92 financially:

• K92 is cash positive operationally as from start of commercial production, but was cash negative after investments until 2019.
• As from 2020 the company is cash positive even after investments.
• With the strong cash flow the company no longer needs to rely on equity financing and has been able to substantially add to the cash balance which reached almost US$35 million by 30 June 2020.

Valuation Of The Kainantu Project

Background

The technical information in his report has been drawn from a NI. 43-101 compliant technical report by Nolidan Mineral Consultants(“Nolidan”), H&S Consultants (“H&SC”), Australian Mine Design and Development (“AMDD”) and Mincore Pty Limited (“Micore”), dated 27 July 2020, with the findings of an updated preliminary economic assessment (“PEA”) study for the project. Unless specifically otherwise stated all text, information and illustrations were drawn from this document.

The Kainantu Project is located in Papua New Guinea approximately 180 km west-northwest of the port town Lae (see Figure 3.1_1).

The property is serviced by a 10 km long access road from the Lae- Madang Highway, commencing at Gusap Airstrip to the Kumian Process Plant and Office facility.

The tenement area covers 856 km2, comprising an exploration licence application (201 km2), four exploration licences, one mining licence (ML150 shown in red on Figure 3.1_2), two mining easements and one licence for mining purposes (all shown in orange).

There are no royalty obligations to third parties. The government imposes a royalty of 2.0% of the free-on-board (“FOB”) value of product that was not melted, or 2% net smelter royalty (“NSR”) when the product has been smelted in Papua New Guinea. In addition there is a levy of 0.5% on the gross revenue.

The government of Papua New Guinea did not exercise its right to acquire at sunk cost a 30% in the mine and has no similar rights under the mining licence renewal process.

Geology and Mineralisation

According to the technical report within the Kainantu project area are deposits ranging from Au epithermal gold veins (e.g. Irumafimpa), Intrusion Related Gold Copper (“IRGC”) veins with Au-Cu-Ag (e.g. Kora), porphyry Cu-Au systems and Cu, Pb, Zn mineralisation skarn

Skarn Explainer

“Skarn”-type deposits are formed in a process similar to that of porphyry orebodies. Skarn deposits are developed due to replacement, alteration, and contact metasomatism of the surrounding country rocks by ore-bearing hydrothermal solution adjacent to mafic, ultramafic, felsic, or granitic-intrusive body. It is most often developed at the contact of intrusive plutons and carbonate country rocks. The latter are converted to marbles, calc-silicate hornfels by contact metamorphic effects. The mineralisation can occur in mafic volcanic and ultramafic flows or other intrusive rocks. There are many significant world-class economic skarn deposits: Pine Creek tungsten, California; Twin Buttes copper, Arizona, and Bingham Canyon copper, Utah, United States, OK Tedi gold–copper, Papua New Guinea; Avebury nickel, Tasmania; and Tosam Tin–Copper, India (reconnaissance stage). - S.K. Haldar et al., in Mineral Exploration, 2013

Figure 3.2_1 shows the location of various deposits in the project area. The Irumafimpa deposit has historically been mined and has some mineral resources remaining, but mining is currently taking place at Kora North. For the purpose of this report the Kora, Kora North and Eutompi (adjacent to Kora) deposits are most relevant.

These deposits occur in a zone that is approximately 2.5 km long by 60 m wide in which individual veins vary from less than one metre wide that pinch and swell over short distances (Au lodes) to more continuous veins up to several metres wide (Au, Cu) – rich quartz and sulphide lodes.

The Kora-Irumafimpa vein system follows a number of NW trending shear zones. Veins are breccia veins with abundant clasts of both altered wall rock and earlier stages of vein mineralisation. At Kora both the sulphide-rich Cu-dominant and quartz-rich Au-dominant mineralisation occur along the same NW trending sub-vertical structure. The quartz-rich Au-dominant mineralisation shows modest variations in dip (from sub-vertical to locally -60° dip) and strike. Within these lodes several high-grade shoots have been identified referred to as the Kora, Kora North, Eutompi and Irumafimpa deposits.

The vertical extent over which the shoots have been delineated is significant, with Kora identified outcropping at the surface at 1,950 m above mean sea level (“amsl”) with drilling having confirmed a vertical extent greater than 1.2 km, and still open at depth. The Kora veins average 3.4 m true width, which is the entire extent of the known veins before cut-off grades are applied.

Irumafimpa outcrops at 1,300 m amsl with the veins averaging 1.2 m true width, which is the minimum width used during resource estimation.

Eutompi is the area of mineralised lode between Kora and Irumafimpa, extending from around 58,900 N to 59,400 N.

Figure 3.2_2 is a longitudinal section of the deposits for which mineral resources have been defined.

The section refers to the Kora-Eutompi-Kora North resource area as “Kora Consolidated”.

The Kora Consolidated deposits consist of a relatively complex structural zone hosted in which the mineralisation comprises narrow high-grade gold quartz-sulphide veins, veinlets and disseminations. At Kora North there is a lode (referred to as K1E) in the footwall of the structural zone. K1E has a marked higher gold grade than other lodes in the general Kora/Irumafimpa area. The K1W lode has been identified immediately adjacent to the hanging wall of the K1E lode and has a much lower grade gold zone.

The K1W lode is structurally overlain by a clay fault zone of varying widths. A distinct separation comprising relatively barren rock then occurs before reaching the hanging wall zone of the structural zone where the K2 lode resides. This lode is noted in the Kora North area for its relative higher copper content compared to K1, occurring as chalcopyrite (CuFeS2) blebs, veins and masses associated with modest amounts of quartz veining.

Mineral Resources and Reserves

Mineral Resources

The database consists of 266 drill holes totalling 60,443 m and 837 underground samples of the working face. The cut-off date for information was 2 April 2020.

The geological model for the 2018 resource estimation used wireframes for the K1E, K1W and K2 lodes, which were based on a 1 g/t gold cut off. However, reconciliation with mill production indicated that using these wireframes considerably understated the amount sent to the mill and the gold ounces produced by the mill, because actual development drives and stope widths were considerably wider.

For the 2020 resource estimation a larger K1 wireframe shape was generated, combining the K1E and K1W lodes, and was based simply on the presence of gold mineralisation derived from the assays i.e. using a much lower, 0.1 g/t - 0.2 g/t Au cut-off grade. It also considers geological control and a nominal minimum mining width of the current stoping and development of approximately 4.5 m wide. A similar process was applied to K2, which also resulted in a larger wireframe. In addition to the drill hole information, the channel samples from development for the test mining area were used to locally guide the interpretation of the mineralised vein wireframes. The geological interpretation extends 75 m along strike and down dip beyond the last mineral intercept.

The new wireframes were primarily constructed as strings on 10 m spaced E-W sections expanding to 25 m spaced sections at the southern end and 20 m spaced sections at the northern end.

Figure 3.3.1_1 is an example of a cross section with the geological interpretation for Kora North using both drilling and face sample data. The image shows the K1 lode (red dash), the K2 lode (green dash) and two of the Kora Link lodes, KL1 (purple dash) and KL2 (light brown dash). It should be noted the veins between K1 and K2 are particularly well developed at this cross section and the illustration should not be considered representative of the deposit.

The illustration shows the logged vein on one side of the drillhole trace along with the gold grade as a scaled coloured bar on the other side. The development drives and stopes are shown as a brown solid outline. The lengths of the colour bars are intended to be relevant to the gold grade, but gold grades are so high that the maximum length has been limited at 30 g/t. Core was sampled 5 m either side of the lode to include possible stringer style mineralisation away from these lodes.

A sample composite length of 1.0 m was selected, which is also the most dominant sample interval, with a minimum composite length of 0.5 m to fit within the wireframe outline. Sample values within a designated lode were only used to inform block grade values within that wireframe, called “using hard boundaries” in the industry.

Figure 3.3.1_2 has been included to give an impression on the distribution of composite data and the variability in gold grade for the two most important lodes: K1 and K2.

For K1 the clustering and consistent nature of the grade stands out in the dominant N-S flat-lying area, which is the current stope development area at Kora North. Away from here the density of information is much reduced and grades are less consistent. The area to the right (= north) of the sample cluster has conspicuously lower grade. Not shown in the illustration is the same longitudinal section for K1 but with copper grade composites. These show clearly higher values in the top left (Kora), which are much more consistent than for Kora North and Eutompi (at the top right of the section).

The longitudinal section for K2 shows a more consistent grade pattern, but with lower gold composite grades for Eutompi.

Based on the experience that the mean gold grade for underground samples exceeds the mean for diamond core samples, H&SC prefers not to apply top-cuts and instead “control[s] any potential higher-grade outliers through judicious use of the composite interval, grade interpolation parameters, block size and the geological interpretation”. Whatever this exactly means in practice is not clear, and as we discuss below, there is risk associated with not applying top-cuts.

The summary statistics for the K1 Lode shows a coefficient of variation (“CV”) of 3.7 for K1, which H&SC rates low, and 6.1 for K2 after cutting two very high-grade composites to 1,000 g/t Au. Crux Investor suspects the lower CV for K1 is partially explained by the more consistent grade of the sample cluster than for K2. The CV of the sample population may well increase substantially away from this area.

The general conclusion of variography analysis (which measures the degree by which the grade variation increases with distance between data points), is that the results are weak, which is blamed on the scarcity of data points away from the development area. The main implication from the variography is that more infill drilling is required.

A block size of 5 m (north) x 5 m (vertical) x 1 m (perpendicular) was chosen and grade interpolation undertaken by ordinary kriging. Because of subtle variations in the dip and strike of both K1 and K2 eight areas were identified for each lode where the search direction was different.

Whereas H&SC estimated mineral resources for various Au Eq grades, it elected to report mineral resources at a cut-off grade of 1 g/t Au Eq, which is very low for underground mining, implying at its assumed gold price of U$1,400/oz cash operating cost of less than US$45/t milled.

Table 3.3.1_1 shows the estimated resources for the Kainantu deposits using a cut-off grade of 1.0 g/t Au Eq.

The table shows that the K2 lode is the most important contributor to metal in mineral resources, accounting for almost 57% with the Kora Link lodes insignificant contributing just above 2%. It is for this reason this report ignored discussing the Kora Link lodes estimation.

What is also evident from the table is the relative low confidence attached to the mineral resources with Inferred Resources containing 3.5x the amount of metal in the Measured and Indicated (“M&I) category. Not shown in the table is that, of the total 4.7 million ounces (“Moz”) Au Eq., Measured Resources only account for 0.3 Moz, or slightly more than 6%.

Figure 3.3.1_1 shows the location of the various resource categories on a longitudinal section to demonstrate the preponderance of Inferred Resources in blue away from underground development.

What is also not directly evident from Table 3.3.1_1 is the increasing contribution of by-product metals copper and silver away from where mining is currently taking place, judging from the ratio Au Eq. grade divided by Au grade. For M&I Resources of the K1 vein this is 1.06, but increasing to 1.21 for Inferred resources. These ratios are high for K2 at respectively 1.12 and 1.25, probably reflecting the much higher copper grade for the Kora area.

In Figure 3.3.1_4 are two cross sections through the gold block model for K1, which are only 60 m apart and have been included in this report to demonstrate the rapid changes in block grades along strike.

Earlier reference was made to H&SC using other measures than top-cutting of grade to limit the impact of very high values. Based on the summary statistics for the block grade, their measures (probably mostly the result of using a high minimum number of composites to calculate the block grade) did have a major impact on the CV which dropped from 3.68 for K1 composites to 1.97 for block grades and from 6.1 for K2 composites to 2.84, which is much better, but still high.

Given the above it is not surprising that the impact of top-cutting gold composite data for K1 has a modest impact of less than 5% in gold ounces. However, the top-cutting at K2 does have a major effect, and remember that 57% of the resource considered as mineable inventory for the expansion plan.

Quoting directly from the PEA,

“The impact of applying the 1000 g/t top-cut for gold is significant as it affected only four samples, but produced a 22% drop in the overall mean value of the composites.”

And

“The effect of the gold 1000 g/t top cut for the K2 lode is significant with just under 700,000 of gold ounces being associated with the four samples >1000 g/t. It should be noted that these samples are outside the face sampling zone, in areas of wide spaced drilling, where the impact of such grades can have a major influence.”

So, no top-cut was applied to the resource as a whole, but in the K2 lode where most (57%) of the resources are sitting, a top-cut applied to just four samples saw a 22% or 700,000 oz drop in resource. Putting it mildly, we conclude that given the large Inferred component to total mineral resources and the impact of a few samples for K2 on total metal content (accounting for 15%), there is considerable risk associated with the resource estimation. This is however not unusual for underground mines at it is too expensive and impractical to drill out the deposit to a degree that it would cover many years of production. Except for some exceptional cases, underground mining operation need a planning time horizon of around 3 years to be able to design underground infrastructure and access stoping blocks to sustain steady state production.

As the technical codes prescribe that mineral reserves can only be declared using M&I mineral resources, it is not surprising K92 Mining prefers to issue PEA’s reports that do not impose such a restriction and allow for inclusion of Inferred resources in its production schedule. To distinguish these from “mineral reserves”, which is technically forbidden, they are often referred to as “Mineable Inventory”.

Mineable Inventory

The estimation of mineable inventory makes a number of very important assumptions, being:

• The dominant mining method will be longhole open stoping contributing 80% to the plant feed with the balance by cut-and-fill mining, which is the current mining method. The desire to change mining method is the higher productivity of longhole stoping and management planning production ramp up from the current 0.2 Mtpa to 0.4 Mtpa in 2021 and 2022, to 1.0 Mtpa from 2024 onwards. It is virtually impossible for cut-and-fill to be able to sustain such a rate
• The change in mining technique must be seen as totally aspirational as insufficient geotechnical information is available to support longhole open stoping. Mining experience has demonstrated variable ground conditions for the K1 and K2 veins ranging from poor to good. K1 has poorer ground conditions compared to K2, mainly due to a fault, or clay gouge structure on the western wall (hanging wall) of the K1 vein. In 2020 longhole open stoping was carried out on a trial basis and was found successful, subject to “specific measures to address the hanging wall gouge”. However, the “clay gouge zone is currently not well defined or modelled for application in preliminary stope design”. Clay or Fault gouges are clay-rich zones formed by a combination of intense hydrothermal and/or tectonic activity, and they are structurally very weak, are often a conduit for water flow and can cause significant ground condition and mining challenges.
• Whereas the geotechnical consultants have preliminary suggested conditions for longhole stoping, it has proposed a programme to collect “suitable data and analysis for the upcoming feasibility study mine planning”.
• The preliminary suggested design parameters include a 20 m vertical level spacing for an effective stope height of 25 m, and assume a skin of competent material will be left against the clay gouge in the hanging wall, for dilution these assume a skin on either side of the vein of 0.5 m (total 1.0 m dilution) with the grades of theskin derived from the block model. Further unplanned dilution at nil grade was assumed to be 8% for K2 and 12% for K1. Given the average deposit width of 5.9 m for K1, the total dilution there would be 31% and 26% for K2 with its average deposit width of 6.1 m.
• A mining recovery factor of 90% was assumed.

Instead of using a standard cut-off grade for mine planning purposes, the consultants generated stope shapes at a range of Au Eq cut-off grades from 3.0 g/t to 6.0 g/t in 0.5 g/t increments. A simple production schedule was prepared to provide an estimated pre-tax cashflow and pre-tax discounted cashflow for each cut-off grade scenario and the scenario yielding the higher discounted cash flow vale was selected as having the optimal cut-off grade. This proved to be 5.5 g/t Au Eq.

The stope design parameters of sublevel spacing of 20 m, stope length of 10 m to 20 m, average width of 7 m and density of 2.8 indicate each stope containing on average 6,000 tonnes. Should K92 Mining desire to mine 1 Mtpa it will need numerous stopes developed at any point in time to be able to sustain the target rate. It should also be noted that the consultants have in their design parameters not constrained for the gouge as “this is currently not defined”.

To arrive at a preliminary mineable inventory, a design process was carried out, to ensure that all of the tonnes allocated to inventory are safe to access and make money. Accordingly, after an initial mine design is generated, an optimisation (both in terms of schedule and cost) process occurs leaving a ‘mineable inventory’. From a starting point of stopes with a certain grade and tonnage, a mine design for 2021 and beyond whittles away at stopes that are sub-economic, or which will be consumed in 2020, or which are required for the crown pillar, leaving just the economic stopes as mineable inventory. This mineable inventory is then further adjusted to accommodate unplanned dilution and a 90% mining recovery.

Table 3.3.2_1 gives the preliminary mineable inventory estimated for the Kainantu project.

The conversion rate of the gold equivalent metal cannot be exactly determined as there is no resource definition at the same 5.5 g/t Au Eq cut-off grade used for Mineable Inventory. However, when taking the straight average of resources with cut-off grades 5 g/t and 6 g/t Au Eq, the indication is that all contained metal in mineral resources is included at a grade that is 37% lower than the resource grade, but with 63% more material. This does not seem to reconcile with material lost in a crown pillar, 10% mining losses, etc...

The consultants however make it more complex, by adding as an afterthought low-grade material mineable during the last few years based on using a 3.04 g/t Au Eq. cut-off grade, the location of which is identified in Figure 3.3.2_1.

It should be noted for these low-grade stopes the consultants assume much lower dilution of 15%, the motivation for which is not given. The total low-grade material amounts to 1.58 Mt at an average grade of 3.35 g/t Au Eq. for 0.17 Moz Au Eq.

Mining Operations

The mine plan makes use of existing Irumafimpa development and recent Kora exploration development, and in particular the existing incline, to provide access to the orebody for stope production activities. Currently a twin incline haulage-way is being mined from a new portal north of the existing portal at 841 amsl, which will greatly improve logistical efficiency, benefiting for gravity assisted haulage.

A northern and southern incline will extend upwards from existing underground development levels to the top of the mineable resource. On each level footwall drives will be established to service access of man and material to the stopes.

Figure 3.4_1, which has been extracted from a slide of the corporate presentation dated September 2020, shows the planned main underground infrastructure with the existing development at Irumafimpa and Kora in white, the twin incline and the main ramps and footwall drives.

The cut-and-fill mining stopes at Kainantu have been mined at 2.5 m lifts, which limits the surface area of any working face prepared for blasting. This method therefore imposes severe limitation on the amount of ore than can be mined. Figure 3.4_2 shows the proposed longhole stoping mining method to show schematically the much larger blocks of ore than can be blasted compared to cut-and-fill making the method much more productive. The mine is in the process of being reconfigured to longhole open stoping, and reaching the 2021 target of 140,000 ozpa depends on a sufficient number of longhole open stopes being developed to get the requisite tonnage out of the mine.

The company reports that sufficient trial mining has taken place to prove that longhole open stoping is possible. Crux Investor notes that K92 does not yet have sufficient geotechnical work done yet to be really sure that the method will indeed work. There is risk in the assumption.

The illustration also shows that the stope is progressively backfilled to enhance ground stability and to provide a working base for the next stope above. The method limits the strike length of the open stope to maintain stability of the side walls and backs. Stopes will be progressively backfilled from the opposite end to where the stope is being blasted and loaded out as shown in Figure 3.4_2. Again, Crux Investor notes that there is a calculated risk being taken by K92 here. The Company says sufficient work has been done, Crux Investor knows that trial mining gets more supervisory and management attention than usual and that the move to commercial-scale operations on a new method brings risk with it.

Where access is not possible to both ends of the stope, the filling will be done from above with access from the same direction as the mucking in the level below. This method is not preferred as it has a lower productivity rate and higher filling cost.

At the moment the backfill material is development waste, but the Company intends to use paste fill by 2023. Paste fill is a mixture of processing plant tailings and cement. With cemented paste fill, stopes can be mined in a top-down direction as well as bottom-up. The advantage of top down mining in combination with paste fill is the reduced risk of self-propagation within the clay gouge zone and higher flexibility in mine planning. Paste fill is a proven method of filling with a consistent material of known mechanical properties that gives excellent support to walls. Used here it will help to prevent any spontaneous scaling up to the clay gouge zone which would cause extra dilution.

Processing Operations

There is surprisingly little information on metallurgical testwork and plant performance in the various PEA reports. This despite statements like “the inability to inform the plant metallurgists of impending feed characteristics often resulted in dramatic consequences and inefficiencies in the operation of the plant” (Nov 2016 PEA), “the deleterious impact of clays” (Jan 2019 PEA) and “the operational difficulties experienced with the current crushing circuit” (April 2020 PEA). It is therefore not surprising the April 2020 report contains a long list of required test work to support the design of the 1 Mtpa plant.

The suggested plant recovery performance of 95% for both gold and copper and 77% for silver must therefore be seen as aspirational as it exceeds the current 93% gold recovery and 92% copper recovery, whereas more clay content is expected in future.

The planned 1 Mtpa process plant is expected to include a flow sheet of crushing and grinding to P80 minus 75 μm, flash flotation and gravity concentration for recovery of free gold followed by conventional sulphide flotation to a high gold copper concentrate.

Economic Valuation – Kainantu Project

Metal Prices Assumed

For the Base Case of this valuation, the gold spot price on 2 October 2020, US$1,899/oz Au was used to determine the value of the discounted cash flow.

Crux Investor has used the Preliminary Economic Assessment as the basis of the economic valuation of the mine, even though it feels that the PEA is problematic for a number of reasons. A PEA is only a preliminary economic study, and it is therefore understood, and totally natural that all of the numbers have a wide margin of error. But a good PEA covers all of the topics and notes where further information or test work is need. In this PEA, Crux Investor notes that:

• Whereas copper is supposed to become a more important by product over time, the PEA production schedule does not provide detail on feed grade, metallurgical recovery and concentrate grade over time to be able to generate the required information to model copper’s contribution. Instead the production schedule gives Au Eq. grade, locking in the by-product contribution irrespective of actual metal spot prices. This valuation had to hard code the 15% contribution of by-products to allow for sensitivity analysis with respect to the gold price.
• There is a general lack of detail which makes it impossible to reconcile numbers and to model annual cost figures.
• There are obvious oversights in the PEA model with cost items ignored.
• The suggested cost structure bears little relation to current reported costs.
• The model is on a pre-tax basis ignoring the largest cash drain, which even exceeds PEA operating expenses and for this valuation is the second highest cash drain.

Acknowledging these shortfalls in the PEA, this valuation will therefore present the post-tax results for the PEA scenario and a scenario for spot prices and amendments to the cost structure using benchmark, best-guess, or industry standard costs.

Production Schedule

The production schedule is best presented by a number of graphs extracted from the PEA technical report. Figure 3.6.2_1 shows at the top the amount of waste material from development that needs mining and amount of mineralised material produced by source.

The two graphs above show that the development of the mine has to be front-loaded in years 2021-2026 (top graph) in order to support the ramp up in ore production to enable steady-state production of 1 Mtpa from 2026 (bottom graph). This is entirely logical if one thinks about it. In order for a large tonnage to be produced on a steady-state basis, a mine needs lots of stopes operational, and so development tonnes need to proceed production tonnes.

With development dropping as a proportion of stope production insufficient waste will be available for backfilling, which is another reason for requiring paste fill (on top of the superior mechanical properties that paste fill offers).

Peak total mine production is reached in 2025 and 2026 with 1.8 Mtpa ore and waste moved each year. As the consultants observe, there is considerable risk and uncertainty associated with the production forecast. The main factors are the high proportion of Inferred Resources (81%) included in the schedule, uncertainty around rock conditions, in particular the effect of clay gouge, hydrogeological conditions and mining productivity.

This valuation has taken the production schedule at face value because although there is risk associated with it, it is also achievable.

Capital Expenditure

The PEA report does not present a summary table for capital expenditure provisions, but instead gives tables for the most important capital items only and without a life of mine schedule for mine development expenditure, which at US$269 million is by far the most important item.

This leads to contradictory numbers like US$14.3 million for “Tailings Storage Facility Lifts” in Table 67 of the PEA report and US$14.1 million in Table 78 for “Tailings Management Facility” (which is almost exclusively the backfill plant) before Owners Cost and Contingency and US$19.8 million thereafter. These details and inconsistencies are signs of a lack of thoroughness in the report compilation. These discrepancies are important. US$19.8 million is 40%, or US$5.7 million more than US$14.1 million, and in the table below there is no line for Contingency or Owners Costs. Without contingency shown separately the tailings management facility and paste fill plant in the table below (Table 3.6.3_1) should be US$19.8 million, not US$14.1 million. Such inconsistencies make one wonder how rigorous the total estimate was put together.

Under the Economic Analysis section of the PEA report there is a statement that upfront capital expenditure is estimated at US$124.6 million and US$341.3 million for LOM sustaining expenses, so a total of US$465.9 million.

Table 3.6.3_1 gives the breakdown of capital expenditure as derived from the detail provided in the PEA report, the total of which is US$466.3 million. The difference between the CapEx figure in the table, and the CapEx figure in the previous paragraph is only US$0.4 million, but it is another inconsistency.

As well as the tailings and paste fill undershoot by US$5.7 million, there are a couple of other numbers that seem too low. To expect light vehicles to only require US$0.7 million over the LOM is not realistic, in our opinion. Furthermore we believe that a process plant expenditure of just over US$46.3 million for a 1 Mtpa facility in an isolated and geographically difficult area is too optimistic.

The provisions do not include an amount for the feasibility study, or the substantial drilling that is required to upgrade the mineral resource to an acceptable confidence level for mine planning purposes.

However, with underground development and other capital items relating to mining accounting for almost 80% of the total, most attention should be focused here. Referring to the risk associated with the production schedule, this risk is transferred to capital provision for mining activities. Even so, this valuation has adopted the capital estimates and will under the sensitivity analysis section determine what is the financial risk associated with capital expenditure.

Operating Expenditure

Table 3.6.4_1 compares the cost structure suggested by the consultants for the Kainantu project at a plant throughput rate of 1.0 Mtpa against the actual cost structure over the last 2½ years to check how realistic is the forecast.

Once again, the consultants do not give a full summary table for total site expenditure. What is obvious is they overlooked the paste fill plant operating cost in their cost structure. Moreover, they included concentrate transport and treatment and refining charges as part of the processing plant cost instead of accounting for this as off-mine charges. The consultants cost rate for processing of US$25.20/t treated includes the off-mine charges and US$1.0/t for sustaining capital expenditure. It is unusual to include concentrate transport and treatment and refining charges as part of the processing plant cost because processing plant costs are driven by the processing costs per tonne of feed, where the downstream costs are driven by the transport and treatment and refining cots per tonne of concentrate. Lumping the two sections together muddies the waters and makes it harder to analyse.

The table above shows how low the forecast total site operating cost compares to current actual costs. Crux Investor does not feel that the future estimates of cost are credible in the light of actual expenditure and known current costs, even when considering benefits from the economies of scale and from longhole stoping in reducing operating expenditure. In theory longhole stoping will be significantly cheaper than the current cut-and-fill method, but there is considerable technical risk to overcome and for the mining method to prove that it is suitable for these ground conditions. Further, the need to buy, commission, and operate a paste fill will reduce some of the saving anticipated in the PEA. In addition, there is a clear omission from the operating cost forecast in the substantial and ongoing delineation drilling required for stope design planning.

Crux Investor also questions some of the scale-up factoring that is incorporated into the PEA. Using standard assumptions for fixed/ variable cost split (e.g. 30/70 for processing), is not appropriate for the five-fold increase in production rate as certain “fixed” cost component become substantially variable at quantum changes in production rate. For example, the supervision staff complement for a 0.2 Mtpa will never be able to cope with 1.0 Mtpa production.

Unfortunately this valuation had to make a guesstimate about the LOM cost rate for 1.0 Mtpa, settling on US$170/t which is 83% higher than the average PEA rate, but 44% lower than current unit costs. There is much uncertainty around the number, but Crux Investor feels that it is a more appropriate number to use than the US$93.5/t suggested in the PEA.

In addition, this valuation takes account of corporate overheads as the valuation is of the company and not just its project. Presently the overheads amount to US$2.2 million, but this can be expected to increase substantially with the fivefold expansion in production. A rate of US$4.6 million per annum has been assumed.

Working Capital

The PEA study ignores investment in working capital, which for an operation producing a concentrate and having a long development and construction pipeline before receiving revenue from what it produces is a considerable oversight. At 30 June 2020 the investment in net current assets, ignoring cash, was US$24.6 million which is dominantly receivables as the creditors basically finance the stores and work in process inventory.

Of the US$24.6 million, GST receivable was US$4.4 million and the balance account receivable of sold product. This is approximately 6 weeks revenue in Q2 of 2020. For this reason investments in net current assets were modelled at 6 weeks revenue plus US$4 million as a minimum transactional cash balance.

Royalties and Taxes

With reference to Section 3.1 there is a royalty of 2.0% of the free-on-board (“FOB”) value of the concentrate product plus a levy of 0.5% on the gross revenue.

The PEA has carried out the economic assessment ignoring taxation. No information is provided in the technical report on tax regulations. For this reason reference was made to “Tax Facts and Figures 2018 Papua New Guinea” by PWC and in particular to a note by the Department of Mineral Policy and Geohazards Management entitled Papua New Guinea Mining Fiscal Regime.

Applicable taxes for mining companies in Papua New Guinea are:

• Income Tax for resident companies is 30% (for non-resident 40%, but deemed not applicable)
• Dividend withholding tax is 15% even for countries with which there exist a tax treaty.

Expenditure incurred during the exploration phase of a project (i.e. under an exploration licence) is categorised as exploration expenditure (“EE”). EE incurred up to 20 years prior to the issue of a development licence qualifies as allowable exploration expenditure (“AEE”).

Capital expenditure, including exploration expenditure, incurred after a development licence has been issued is categorised as allowable capital expenditure (“ACE”). Both AEE and ACE can be deducted for income tax purposes, up to the amount of income in that particular year. The AEE deduction cannot be used to create a tax loss.

Without the technical report providing details on ACE and AEE, this valuation has referred to the financial statement for the year ending 31 December 2019 where on page 15 it provides the balances and additions to Property, Plant and Equipment. This was used to determine the annual year-end balance of the ACE pool. It indicates that at the end of the current financial year the company should have redeemed its investments for tax purposes and future tax allowances are equal to the capital expenditure for the year.

Results

Table 3.6.7_1 gives the forecast financial performance for PEA input parameters and the amendments incorporated into this Crux Investor valuation, and at Base Case metal prices.

The calculated pre-tax cash flow of US$2.77 billion for the PEA scenario differs by US$85 million from the quoted number of US$2.86 million, which is explained by the PEA ignoring US$67 million paste fill plant operating cost.

The table shows the extraordinarily high operating margins for both cases: almost 91% for the PEA scenario and 74% for this valuation scenario. It again demonstrates that nothing beats grade for mineral deposit projects. With such high margins the 70% upward adjustment of operating mining cost has relatively little impact, more than compensated by the 27% higher gold price.

Crux Investor feels that it is not appropriate to apply the 5% discount rate customarily used for operating companies as the number of risks associated with the ramp up to 1.0 Mtpa is so extensive, the Kainantu project should be considered a “new project” for which a discount rate of at least 7.5% should be used.

Table 3.6.7_2 expresses the sensitivity of the value of K92 as the change in Net Present Values per percentage point change in the economic main parameters.

The table shows that, for every percentage point increase in gold price (i.e. US$19.0/oz) the NPV7.5 increases by US$27.7 million, which is only 1.6%, and for every percentage point increase in the cash operating cost (i.e. C$1.70/t treated) the NPV7.5 drops by C$9.6 million, which is 0.5%. The sensitivity to capital expenditure changes is even lower with a drop in NPV7.5 of 0.2% resulting from a 1%-point increase.

The Kainantu project is economically extremely robust.

The Enterprise Value of K92 Mining

Figure 4_1 shows the forecast distributable cash flow over the LOM based on the input parameters of this valuation.

The graph illustrates that cash flow is very much a function of being able to achieve the five-fold ramp up in processing rate. This is a risk that is difficult to quantify. Investors should keep this risk in mind and monitor to what extent the planned studies support the assumptions on which the 2020 PEA mine plan is based.

At the share price of C$6.88 on 2 October 2020 and with 213.05 million shares issued, according the TMX money website, the market capitalisation of K92 is C$1,466 million, or US$1,102 million. At 30 June 2020 the company had a total number of 16.81 million share options outstanding of which all are in the money (the highest exercise price is below C$4.99).

The non-cash net current assets at 30 June 2020 have been ignored as their realisation at the end of LOM has been accounted for in the cash flow model. On that date the company had a cash balance of US$34.7 million of which US$4 million has been assumed as part of net current assets as the minimum transactional balance.

The loan balance on that date was US$9.1 million.

Based on the above a diluted Enterprise Value for K92 of C$1,528 million (US$1,149 million) is derived as shown in Table 4_1.

Compared to the NPV7.5 the Enterprise Value is at a 34% discount. Put another way, it is trading at 0.66x NPV. It does not often happen that the market value is lower than the “intrinsic value”, which could be a reflection of the uncertainty surrounding the mineral resources and the very high risks associated with the project, as well as the discount that will always apply to a single asset company.

A single asset company in a frontier market will trade at a discount to full value due to the risk associated with it not having a diversified cash flow stream. The revenue stream from a single asset is binary in that it is either flowing, or it is not. Strikes, pestilence, plague, earthquakes, revolution can all happen (Covid-19 anyone?), let alone the commonplace ground failure or labour dispute. For a diversified company the impact of one mine closure can be mitigated by support from other divisions. Not so for K92, with all of its eggs in the Kainantu basket, with everything hinging on the safe expansion.

Crux Investor feels that as Kainantu matures as a project, management will come under pressure to diversify the asset base, and to mitigate risk of being a single asset company. For the time being, however, the low cost of resource ounce discovery, the geological endowment of the mining lease, the high grade asset all mean that any other project will struggle to match the return of investing in Kainantu itself. There is a lot of runway at Kainantu in the coming years.

The EV per “mineable inventory” ounce figure is US$353/oz, which is probably about right for a single-asset producer. The high EV/oz figure reflects both the high grade and high margin production, and also the prospect for significant resource growth. Crux Investor believes that the quality of the asset, the mineral endowment and the potential for spectacular discovery justify this valuation.

Even accounting for a single-asset discount, the valuation of K92 at an EV/NPV of 0.66x is, in our opinion, harsh. Crux Investor feels that, provided the performance over the next year is in line with the PEA plan, the discount will be unlocked, and a reasonable level would be getting towards 1.0x of NPV, which would suggest a 52% rise in the current share price.

Note that the Tier 1 gold producers are currently trading at around 1.3x NAV, and that the NPV figures calculated by Crux Investor do not include the option value of future resource discoveries which are discussed in the next section.

Blue Sky Potential For K92 Mining

Figure 3.2_1 in this report identified the numerous prospects in the vicinity of the current mine. Based on the September 2020 corporate presentation, K92 management attach the highest priority to the Judd vein and Karempe veins which run respectively east and west of and parallel to Irumafimpa-Kora vein system. Drill results shown on presentation slides are partially spectacular, but also very inconsistent. It is however very early days and the veins systems are definitely very prospective.

Whereas Judd and Karempe are potential narrow high-grade deposits, the other highlighted prospect, Blue Lake, has the potential to be an Au-Cu porphyry.

Figure 5_1, which is a longitudinal section looking southwest through the Blue Lake target showing boreholes with subeconomic grades of around 0.2% Cu and 0.2 g/t Au, but consistently ending in mineralisation with K92 holding out the prospect of higher grades towards the core of the target.

The exploration results are too uncertain to put a value on. They just add blue-sky potential to the company.

Given the obvious considerable size of the mineralised system in the Kainantu project area, evident from the numerous targets in Figure 3.2_1, it is highly unlikely the company will not find additional resources in due course.

Red Flags

At risk of repeating the prior chapters, Kainantu is a fascinating prospect but it is not without its challenges. The mine has consistently returned more gold than was expected from ore, but the PEA study contained some omissions and flaws. The management team has done an excellent job on the operation so far, but now the Company is undergoing a major expansion on a single asset with resource estimate risk, and perhaps most importantly a new mining method on relatively narrow veins and some difficult ground conditions. The Ying and the Yang.

When trying to ascertain the balance of risk and reward, challenge and opportunity, Crux Investor comes back to the key facts that the asset is great. High grade, under-explored, highly prospective, and producing strong cash flows, Kainantu is the real deal.

As ever, the best way to review a company is to strip away the emotion, and look at the hard numbers. When Crux Investor puts hard numbers on top of the partial data provided by the PEA, it is noteworthy that the NPV generated is a 34% premium to the EV figure. On top of that is the exploration potential of the full licence, not quantified here. And a further key point is that the gold sector as a whole is trading on a ratio of 1.3x NAV, so K92 is trading on a relatively punitive basis.

Still, for completeness’ sake, here is the list of Red Flags and Green Lights...

• Single asset company, meaning that the revenue stream is either flowing or it is not. Binary outcome for the bottom line, which will ensure that the Company will always trade at a deeper discount than diversified peers
• PEA contains some inconsistencies, unrealistic (in our opinion) costs, missed some essential expenditures, and incorporates some risk
• High proportion of Inferred Resources relative to Measured and Indicated Resources
• Lack of top-cuts in the estimation, when by the PEA’s own admission a top-cut would reduce the main resource lode (K2) by 22%, or 700,000 oz
• Switch to longhole open stoping carries the inherent risk of being a new mining method, despite the switch making eminent sense from an expansion perspective
• Unproven ground conditions and the unknowns associated with the Clay Fault Gouge zone when longhole open stoping
• Expansion plan is ambitious and targets set by mining companies are often missed
• Unrealistic (in our opinion) operating costs used in the PEA, adjusted by Crux Investor for the valuation work
• PEA valuation calculated pre-tax, whereas any real-world analysis needs to be post-tax

Green Lights

• Strong management track record established since operations began in 2016
• Excellent grade, with Kainantu among the highest grade operating mines worldwide (15.6 g/t Au Eq. in H1 2020)
• Low discovery cost per ounce, testament to the mineral endowment of the property
• Multiple exploration opportunities, and resource expansion potential at depth, in the mine area, and further afield within the broader exploration licences
• Very strong cash margins (74%), a function of grade, price, and cost
• Production forecast to grow from 82,000 oz in 2019 to 318,000 in 2023, on a lower cost base per ounce
• K92 trading at a 30% discount to NPV 7.5, and 15% discount to NPV 10
• Peers trading at 1.3x NAV, so relative valuation is punitive

These CRUX Reports are written for expert investors AND for people new to natural resource investing. But whether you are an expert or a newbie, we all have the same driver. We invest to make money. Sometimes investors get emotional about the investment. They actually think they own a mine. They don’t. They own shares in a company. So focus on your investment strategy, work out the best plan for your needs, stick to the fundamentals and remember that the only way you make money is if your shares go up in value…assuming you don’t forget to cash them in!