Endeavour Mining — Entering a new phase

Since our last note on the company (see New senior gold major looking to join FTSE 100, published on 17 December 2020), Endeavour has made six major announcements:

This note seeks to update our estimates and valuation of Endeavour to incorporate all of the above.

On 23 February, Endeavour reported positive PFS results for both its development projects, Fetekro and Kalana. Full details of the announcement are available on Endeavour’s website. However, below is a brief summary of the results of the outcomes of both studies relative to the most recent studies conducted beforehand:

Readers are cautioned in making direct comparison between the economic results of the 2021 PFS for Kalana and its 2016 DFS, given that the prior study was conducted at a gold price of US$1,200/oz and used an 8% discount rate (cf US$1,500/oz and 5% in the updated study, respectively). Otherwise, in the case of Fetekro, it can be seen that the plant size has doubled, tonnes processed over the mine life have more than doubled and gold production has doubled for only an incremental increase in both capex and opex. In the case of Kalana, the plant size has similarly more than doubled; however, it will produce approximately the same amount of gold over the life of the mine, albeit at a higher rate over a shorter period of time, similarly for only an incremental increase in capex and opex.

Detailed mine schedules and costings for each project were provided in Endeavour’s announcement of 25 February and these have now been incorporated into Edison’s financial model. In the meantime, definitive feasibility studies (DFS) on both will commence. In the case of Fetekro, its DFS is anticipated to be completed in Q421. In the case of Kalana, it is expected to be completed in Q122. For the purposes of its financial modelling, Edison has assumed that 2022 will be occupied with final adjustments and mine plan optimisations and financing arrangements before a final investment decision is made later in the year. We then assume that initial capex will be expended in FY23 and FY24, with first production and the last of initial capex to occur in FY25.

In its preliminary announcement of 25 January (full FY20 results were later published on 18 March), Endeavour revealed record Q420 production of 344koz (which represented a 41% increase cf Q320) at an all-in sustaining cost (AISC) of c US$770/oz. The company also reported record consolidated FY20 production of 908koz and a net cash position of c US$70m at year-end (NB this does not match the net cash/debt position shown in our ‘Financial summary’ table – Exhibit 10 – at the end of this report, which we have shown as consolidated with Teranga as at end-FY20). More detail on the financial effects of the Teranga acquisition are provided below.

Full details of the performance of each of Endeavour’s mines is provided in its announcement. In summary however, it is probably sufficient to say that all mines outperformed our expectations in terms of both production and costs for the period with the single exception of Agbaou (which was subsequently sold to Allied Gold).

In financial terms, Endeavour’s revenue for Q420 of US$604.1m was US$54.4m (9.9%) above our prior expectation of US$549.7m (see the columns entitled ‘Q420e’ and ‘Est Q420a’, below). However, this is exactly in line with the additional production and sales recorded during the quarter relative to our forecasts. At first glance, this was counteracted by operating expenses, which were US$71.8m (37.8%) above our forecast of US$189.7m; however, this measure of costs excludes the reversal of the fair value adjustment of inventory on hand at the SEMAFO acquisition date for both Mana and Boungou and also the write down of gold-in-circuit pertaining to historically stacked ore that has now been deemed to be unrecoverable at Karma. If these adjustments are excluded, then operating expenses were only 3.3% above our prior forecast, allowing the US$54.4m in additional revenue to feed through almost directly to a US$53.4m (60.3%) positive variance in adjusted net earnings for the period relative to our prior forecast (US$142.0m vs US$88.6m). Readers should note that, while these cost adjustments are excluded from Endeavour’s income statement, they are included in its calculations of both adjusted net earnings attributable to shareholders and also total cash costs and all-in sustaining costs.

In addition, Agbaou was pre-emptively deconsolidated from Endeavour’s group accounts and held as an asset for sale for both Q420 and FY20. In the following table, we have attempted to reconsolidate Agbaou into Endeavour’s accounts (see columns marked ‘Est Q420a’ and ‘Est FY20a’ in contrast to ‘Reported Q420a’ and ‘Reported FY20a’, respectively), so that a more direct comparison may be made between the actual results as reported for both Q420 and FY20 and previous periods, as well as our prior forecasts (although readers should note that the cost adjustments, which were made for both Q320 and Q420 and which have been alluded to above, have been jointly loaded into Q420 for the purposes of our reconsolidations, below – which accounts for the majority of the difference between Edison’s ‘Est Q420a’ adjusted net earnings attributable number of US$141,985k and Endeavour’s US$163,602k).

One further characteristic of the results was a sharp drop in the depreciation charge, overall, and at Boungou and Mana, in particular. This was something of a surprise, given that Endeavour calculates depreciation on a units of production basis and production in Q420 was materially higher than in Q320 (even in the reported case in which Agbaou was deconsolidated). This apparent anomaly may be explained by the decision to retrospectively recognise goodwill in Endeavour’s acquisition of SEMAFO to be treated separately from the depreciation of the underlying assets (as had been the case in Q320).

Finally, Endeavour changed its definition of cash costs during the course of the quarter to include royalties. The decision was made for Endeavour to be more consistent in reporting within the context of its peer group. However, readers should note that, for reasons of comparability with past results, we (at least for the moment) continue to show our own calculations of total cash costs excluding royalties unless specifically indicated otherwise (eg in Exhibit 2, below). We can confirm that, barring this one difference, our own calculations of total cash costs reconcile with those of Endeavour exactly. Otherwise, readers wishing to see calculations of total cash costs including royalties are directed to Endeavour’s management discussion and analysis (MD&A) and integrated financials.

Items included in the reconciliation between adjusted net earnings attributable and total net and comprehensive earnings are losses from discontinued operations, deferred income tax effects, gains/losses on financial instruments, other expenses, share-based compensation and acquisition costs (all shown independently in the table above), plus the tax impact of adjusting items, non-cash and other adjustments and the minority interest attributable to the adjusting items (not shown independently).

A comparison between Endeavour’s actual Q420 and FY20 adjusted net EPS and analysts’ expectations immediately prior to its results’ announcement is as follows:

On 10 February, Endeavour announced the completion of its acquisition of Teranga Gold to create a new, top 10 global gold producer.

The acquisition had originally been announced on 16 November, with Endeavour confirming that it had entered into a definitive agreement with Teranga, whereby Endeavour was to acquire all of the issued and outstanding securities of Teranga by way of a Plan of Arrangement under the Canada Business Corporations Act. Under the terms of the agreement, Endeavour was to pay 0.47 shares for each Teranga common share, resulting in the issuance of an additional 78.8m shares (cf 163.0m in issue at the time). As with its SEMAFO acquisition in FY20, Endeavour’s principal investor, La Mancha, committed to invest US$200m into the combined entity to provide for a larger free float and greater stock liquidity and both sets of board of directors unanimously approved the transaction.

Edison has now completed its consolidation of Endeavour’s and Teranga’s accounts (see Exhibit 10). For the purposes of financial modelling (and for making valid subsequent comparisons), we have assumed that the two balance sheets were consolidated as at end-December 2020. At the time, we estimated Teranga’s net assets attributable to shareholders to have been US$870.8m and its net debt to have been US$242.6m. As at 10 February (the date on which the acquisition closed), Endeavour’s share price was US$21.34 making its bid worth US$1,680.5m in equity and resulting in the creation therefore of US$809.7m in goodwill at the moment of consolidation (Edison estimate).

Finally, on 25 February, Endeavour announced updated guidance for the company for FY21, including the assets of the former Teranga. Full details of the updated guidance are available on Endeavour’s website. In summary, guidance for its other six mines was left unchanged in terms of both production and costs. However, guidance of the following was added for Teranga’s assets:

Readers should note that this guidance is consolidated in that it includes Sabodala-Massawa and Wahgnion guidance from 10 February onwards. For the full-year (which forms the basis of our pro forma estimates, below), we estimate production from Sabodala-Massawa of 349.1koz and from Wahgnion of 166.8koz. Production from Sabodala-Massawa is also notable for the fact that its head grade is expected to increase materially in H221 as higher grades are mined at the Sofia pits. In addition, the Phase 1 upgrades scheduled to occur throughout the year will assist in debottlenecking the plant and thereby improve its production capability and efficiencies for processing Massawa’s high-grade oxide ore. Otherwise, the effect of COVID-19 on Endeavour’s operations in West Africa has proven to be negligible to date and is expected to remain so. Nevertheless, Endeavour has mitigated future risks as far as possible by both setting itself up to operate under Level 2 COVID-19 restrictions (see our note, New senior gold major looking to join FTSE 100, published on 17 December 2020) and also by preparing multiple different levels in its pits from which to produce.

Readers should note that, for the purposes of our forecasts, below, we have left Agbaou fully consolidated into Endeavour’s accounts. However, we have also provided our forecast of its individual income statement as an adjunct to our group Q121 forecasts (in Exhibit 5, below), for those who wish to individually deconsolidate it. In the meantime, we understand that it is not Endeavour’s intention, at least for the time being, to reflect Karma as an asset held for sale (despite the fact that it has now been classified as ‘non-core’).

As before, items included in the reconciliation between adjusted net earnings attributable and total net and comprehensive earnings are losses from discontinued operations, deferred income tax effects, gains/losses on financial instruments, other expenses, share-based compensation and acquisition costs (all shown independently in the table above), plus the tax impact of adjusting items, non-cash and other adjustments and the minority interest attributable to the adjusting items (not shown independently).

Within this context, readers should note that one risk to our forecasts (above) arises from potential acquisition costs regarding the Teranga transaction. These we have not attempted to forecast but could be of the same order of magnitude in FY21 as the costs relating to the SEMAFO transaction earlier in FY20 (see Exhibit 2). Such costs will not affect our forecasts for adjusted net earnings per share attributable to shareholders, but could, inevitably, affect our basic EPS forecasts.

Notwithstanding the detailed appearance of our forecasts, readers are cautioned that forecasting on a quarterly basis is prone to large variations between actual and forecast numbers. As such, the exhibits both above and below should be regarded as indicative, rather than prescriptive, particularly with respect to individual quarters. With this caveat, a comparison between Edison’s FY21 adjusted net EPS from continuing operations estimates and consensus estimates, by quarter, is as follows:

While we recognise that our FY21 adjusted EPS forecast is low within the context of the analysts’ consensus, anecdotally we believe that the mean realised price used by analysts in generating their full-year earnings forecasts may be in excess of US$1,900/oz (cf Edison’s US$1,737/oz), which may go some way towards explaining this disparity.

In the longer term and based on the results of the Sabodala-Massawa preliminary feasibility study, a separate refractory ore treatment (ROT) plant using bacterial oxidation (BIOX) technology is expected to be installed parallel to the existing plant to create a second processing stream. The ROT design will increase capacity by 1.2Mtpa and allow for the processing of Massawa's refractory ore, which is expected to begin in 2023.

Endeavour announced the sale of Agbaou to Allied Gold on 22 January. The total consideration for its 85% interest amounted to up to US$80m, being US$20m in cash, US$40m in Allied Gold shares (with an option for Endeavour to re-sell the shares back to Allied Gold at the issue price until 31 December 2022) and US$20m contingent upon the gold price exceeding US$1,900/oz in FY21 plus a variable net smelter royalty on ounces produced in excess of Agbaou’s end-December 2019 reserves. As at end-FY20, Agbaou had US$69.7m in cash on its balance sheet; however, this was largely balanced by US$66.2m in payables (with minimal receivables).

In FY20, we estimate that Agbaou (underlying – ie excluding impairments) earnings were negative (not least owing to a very large tax charge in Q420 and FY20 relating, among other things, to withholding taxes), but that it contributed a net US$10.2m in cash flow to the group (on a 100% basis). However, we estimate that the mine also only had two years of formal life ahead of it (being FY21 and FY22), albeit it potentially has four years ahead of it (ie to end-FY24) if all resources are converted into reserves and duly mined.

The consideration of US$80m for its 85% interest in the mine therefore equates to:

For the purposes of Edison’s financial modelling, we have assumed that Endeavour will receive US$20m in cash in FY21 and US$40m in cash in FY22. We have not, at this stage, ascribed any value to the contingent payments in the event that the gold price exceeds US$1,900/oz or to its net smelter royalty.

Endeavour is a multi-asset company that has shown a willingness and desire to trade assets to maintain production, reduce costs and maximise returns to shareholders (eg the sale of Youga in FY16, Nzema in FY17 and Tabakoto in FY18 and the acquisition of SEMAFO in FY20). Historically, rather than our customary method of discounting maximum potential dividends over the life of operations back to FY21, in the case of Endeavour, we have instead opted to discount potential cash flows back to the start of FY21 and then to apply an ex-growth terminal multiple of 10x (consistent with using a standardised discount rate of 10%) to forecast cash flows in (most recently) FY22. In the normal course of events, exploration expenditure would have been excluded from such a calculation on the basis that it is an investment. In the case of Endeavour, however, it was included in our estimate of FY22 cash flows on the grounds that it was a critical component of ongoing business performance in its ability to continually expand and extend the lives of its mines.

However, Endeavour’s acquisition of Teranga as well as its advancement of Fetekro and Kalana, in particular, has rendered this method of deriving an absolute valuation of the company unreliable. Where before all of Endeavour’s mines were approximately in-phase in the sense that they were in ‘harvest’ mode, its acquisition of Teranga has introduced Sabodala-Massawa into its portfolio, with the potential for US$219m in capex relating to Phase 2 of its ROT treatment plant from 2022, while Fetekro and Kalana potentially add a further combined US$635m in capex in the period FY23–25. As such, an effective ‘second distinct wave’ of capex has been added to Endeavour’s cash flow profile, albeit, with the advantage that this will generate additional free cash flow in the period from FY25 and beyond. Apart from La Mancha’s injection of US$200m in equity into the combined Endeavour-Teranga entity (anticipated imminently), and given that it otherwise has immaterial levels of net debt, there is no reason why this capex should not be funded internally.

Suffice it to say that in the absence of all of these factors, our valuation would otherwise remain very little changed relative to our previous one of US$45.32/share at the start of FY21 plus US$1.08/share (fully diluted) for Fetekro, as depicted in the graph below:

If we were to apply exactly the same methodology to our valuation of Endeavour in the aftermath of its acquisition of Teranga and development of Fetekro and Kalana, we would discount cash flows from FY21–26 and then apply an ex-growth terminal multiple of 10x (consistent with using a standardised discount rate of 10%) to forecast cash flows in FY26. In this case, our terminal valuation of the company at end-FY26 would be US$39.92/share, which (in conjunction with forecast intervening cash flows) would discount back to a valuation of US$35.98/share at the start of FY21, as shown in the graph below:

While this absolute valuation of US$35.98/share in FY21 is 22.5% lower than our previous valuation (35.98/(45.32+1.08)), it is nevertheless still materially above Endeavour’s share price of C$23.92, or US$19.09. Within this context, however, it should be noted that the decline in valuation arises only as a consequence of the assumption of additional (discretionary) capex in pre-financing cash flows in the period FY21–25 and the baleful effect of discounting on more chronologically distant cash flows – especially at a rate as high as 10%.

However, given its elevation into the ranks of the world’s foremost producers of gold, Endeavour can increasingly attract lower cost finance and, as such, a CAPM-derived WACC can also be considered (as we also discussed in our February 2021 initiation of Newmont Corporation). Long-term nominal equity returns have been 9% and 30-year break-evens are currently expecting 2.2% inflation. These two measures alone would imply a real equity return of 6.7% (1.09/1.022) and applying this to our forecast cash flows would imply a terminal valuation for Endeavour of US$59.99/`share and a current valuation of US$55.70/share. Readers should note that, given its beta of 0.66 (source: Refinitiv, 15 March 2021), even this (real) discount rate of 6.7% is likely to prove conservative.

In the meantime, Endeavour’s valuation remains at a material discount to those of its newly acquired peer group, as shown in Exhibit 9, below.

Endeavour’s valuation on a series of commonly used measures, relative to a selection of gold mining majors (the ranks of which it has now joined since its takeovers of SEMAFO and Teranga have been completed), is as follows:

Of note is the fact that Endeavour’s valuation is materially cheaper than the averages of the majors on at least seven of the nine measures shown in Exhibit 9, regardless of whether Edison or consensus forecasts are used. On an individual basis, it is cheaper than the majors on at least 38 out of 54 (70%) of valuation measures, regardless of whether Edison or consensus forecasts are used. Reverse engineered, the average valuation measures of its peers imply an average share price for Endeavour of US$27.00, or C$33.80 per share.

Readers should note that Edison’s forecast dividend yield in year 1 and year 2 (FY22 and FY23) is notional. When it declared its maiden dividend of US$0.37/share for FY20 in November, Endeavour announced a policy of declaring future dividends on a semi-annual basis with the aim of maintaining an approximate dividend yield of 1.6% until it has reached a targeted net cash position of c US$250m (note: by the end of FY21 according to Edison’s updated estimates). Thereupon, it will re-assess its capital allocation priorities, which may include augmenting its shareholder return programme. Since November, Endeavour’s share price has fallen, implying, all other things being equal, that its FY21 dividend should also fall in order to maintain a yield of 1.6%. This we believe to be unlikely and that a held dividend is a much more likely scenario in FY21, or a higher dividend in the event that its share price appreciates above US$23.13, or C$28.97 (being US$0.37/0.016). Given that we have no more information than this however, our dividend ‘forecast’ for FY22 and FY23 therefore shows the maximum that we estimate Endeavour could distribute in order to retain a net cash position of US$250m on its balance sheet. This we believe to be equally unlikely in practice. However, it does indicate that, all other things being equal, Endeavour has plenty of scope to increase dividend distributions to shareholders into the foreseeable future.

In tandem with its results, on 18 March, Endeavour announced a normal course issuer bid (NCIB), or share buyback programme in order to supplement its policy of augmenting shareholder returns. The NCIB commenced on 22 March and will end on 21 March 2022 and will allow Endeavour to buy up to 12.2m ordinary shares, or approximately 5% of its total issued and outstanding ordinary shares at the time of the announcement, whereupon the purchased shares will be cancelled. At Endeavour’s current share price of C$23.92 (US$19.09), the NCIB is worth c US$232.4m and compares extremely favourably with its FY20 dividend payout of US$60.3m and its forecast US$93.4m payout in FY21. Combined, the NCIB and FY21e dividend distribution together represent c US$325.8m in aggregate returns to shareholders – equivalent to a dividend yield of 6.8% – in FY21. Note, however that, owing to the inherent uncertainty surrounding whether purchases are made and at what price under the NCIB, Edison has not attempted to include share buybacks in its financial forecasts in Exhibit 10, below.

According to its results announcement, Endeavour had net debt on its balance sheet of US$43.3m as at end-FY20. However, this excluded lease liabilities of US$37.2m and included an option premium of US$74.6m. Including the former, but excluding the latter results in a net debt position of US$5.8m – immaterial within the context of the company’s balance sheet equity of US$2,057.0m at the same point in time.

As from 10 February, however, Endeavour will consolidate Teranga into its accounts. For the purposes of our financial modelling (see Exhibit 10, below), we have assumed that this will occur retrospectively on 31 December 2020. In this case, we estimate that Endeavour will consolidate c US$242.6m in net debt onto its balance sheet and c US$349.2m in gross debt as a consequence of its Teranga acquisition. As such, on a pro forma basis, we estimate that Endeavour would have had US$323.1m in net debt on its balance sheet as at end-FY20 (43.3+37.2+242.6), which we calculate would have equated to a gearing (net debt/equity) ratio of just 8.8% and a leverage (net debt/[net debt+equity]) ratio of 8.1% on the group’s enlarged equity base (note: all stated prior to La Mancha’s injection of US$200m into the enlarged company).