Caledonia Mining — Update 24 November 2016

As stated, Caledonia’s year-to-date production is on target to meet its FY16 production guidance of 50koz of gold produced. In fact, Caledonia achieved its highest ever gold production for Q316, producing 13,428oz Au, surpassing by 4% the previous record held in Q313 of 12,918oz Au. Including October’s production of 4,149oz Au, Caledonia looks to be able to achieve its FY16 target by producing 9,091oz Au over November and December (before accounting for WIP).

Caledonia maintains its production guidance of c 50koz for 2016. However, it has notified the market it expects downward pressure on earnings due to rand strength, one-off charges related to evaluation of a number of investment opportunities and an increase in share-based payments due to Caledonia’s share price increase.

These result in the following additional costs in our model:

Aside from the non-operational cost increases given above, the following revisions have been made to our model:

The model we use to forecast the gold price implicitly assumes a relationship between the total US monetary base, inflation and the gold price. In 2015 there was an (extremely rare) decline in the total US monetary base and (arguably conservative) absence of inflationary pressures. These combined to reduce the base for our longer-term analysis and, therefore, our longer-term numbers.

For further detail on the above method, please refer to page 48 of our October 2016 sector report: Mining overview: Gold and other metals.

As a result of the above cost adjustments and gold price revisions, and a new forex rate of US$1.24/£ (vs US$1.28/£ previously), our combined dividend discount flow (of Blanket’s free cash flow of production post-IP implementation) and DCF of vendor facilitation loan repayments reduces by 18% from £1.77 to £1.46. This is at a 10% discount rate to reflect general equity risk. Forcing our model’s valuation of Caledonia’s shares to its current share price (114p as of 14 November 2016) requires using a discount rate of 16.3%. This relatively low hurdle rate represents, in our view, a marked reduction in the country-risk factor applied to Caledonia’s shares, as well as a significantly more positive view on the company’s ability to deliver its IP on time and on budget, which is now two years completed.

Caledonia’s staff and management continue to implement the 2014 initiated Investment Plan (IP), with all major components either completed (on time) or remaining on track for completion as envisaged. The elements of the IP completed or being completed are:

In addition to these elements of the IP, the company is also developing a decline (internal ramps) to access the AR South and AR Main orebodies ahead of completing the Central Shaft. We consider these additional costs carried in the capex forecasts for implementing the IP.

The reported head grade for Q316 was 3.36g/t Au, in line with Blanket’s budgeted head grade for the quarter of 3.33g/t Au. Caledonia reported 2.94g/t Au for the month of October, which is below-average for Blanket, however grade control practices have already indicated this grade is starting to revert to normal levels. The following exhibits give historical gold grades, recoveries and tonnages for the Blanket mine, which highlight the increased tonnages being milled and a general improvement in gold head grade since the start of 2015.

Blanket’s processing plant recovers roughly half of the gold delivered to it in ore, via very cheap gravity separation methods (a very overlooked though highly desirable attribute of any precious metal orebody); the rest is recovered by conventional cyanidation and electro-winning processes.

Third-quarter recovery was 93.2%, slightly above Q216’s reading of 93.1%. Both are lower than the budgeted 93.5% expected from processing Blanket’s orebodies. The lower than expected recovery factor is due to:

To help rectify the issue surrounding oxygen mixing, an additional, larger, set of cyclones are being installed which is expected to provide a more consistent grind and homogenous sized feed material to the cyanidation plant. Further, a new oxygen plant is to be commissioned mid-2017, which along with the new cyclones should improve recoveries to the planned 93.5%.

Caledonia had cash at end Q316 of US$12.4m. However, as the company guides to increased costs related to forex, share-based payments and exceptional items (due diligence on an investment opportunity), we expect Caledonia to finish FY16 with cash of US$10.7m. This is post-payment of the company’s total US$2.9m dividend (paid quarterly).

Our FY16 cash balance assumes an average gold price of US$1,253/oz to price production.

As production increases as per the IP, and assuming c 60koz is produced in FY17 and peak capex of US$18.0m is realised, we expect Caledonia to be able to build on its cash balance to the order of US$12.5m by end 2017. This large increase demonstrates the leverage Caledonia has in increasing production and revenues while maintaining a stable fixed cost base.

We attended a site visit on 17 to 18 October 2016, visiting the Blanket Mine as well as a satellite exploration project (Mascot/GG). It is clear that the company has moved some considerable way forward implementing its 2014-initiated IP. Underground bottlenecks that restricted haulage to surface via the No. 4 Shaft have been removed, allowing for the record tonnages and gold production that were achieved in both Q216 and Q316. What was clear first-hand is that existing underground infrastructure could not provide the scope for increasing gold production in a material way, and that the Central Shaft development clearly provides the solution to increasing production and reducing costs.

Seeing this development, in person, it is apparent that post-completion, the Blanket Mine will emerge as almost an entirely new, and far larger, mine, retaining the same continuous, highly profitable orebodies and well-managed mining practices that have characterised the company’s ongoing ability to generate cash, even in the worst of gold price environments.

The following are a few photos of our recent visit to the Blanket Mine site.

Two of the main headframes in use at Blanket today are shown in Exhibits 4 and 5 below. Note the relatively small size compared to the Central Shaft temporary headframe shown in Exhibit 6. This temporary headframe is only used to sink the shaft, after which it will be removed and a permanent headframe, some 2.5x the height of this one, will be constructed and used for run-of-mine operations.

As mentioned previously, Exhibit 6 shows the current temporary headframe in use to sink the Central Shaft to its eventual 1,080m depth.

This temporary headframe will be replaced by a far larger (c 2.5x the height of the one shown) headframe, which will be used to haul men, equipment and materially more ore to the processing plant than is currently achievable via No. 4 Shaft.

Exhibit 8 below is the view from standing at the bottom of the Central Shaft at a depth of 326m. An important geological characteristic of Blanket’s gold ore, is that roughly half of the gold contained can be processed using the simplest of all techniques – gravity separation. This is a key reason for why the company has low cash operating costs. The Gemini vibrating Table shown in Exhibit 9 is basically like panning for gold, and moves the gold to the centre of the table, from where it can be removed by plant technicians. The rest of the contained gold is removed from its waste rock via more conventional cyanidation techniques (carbon-in-leach is the specific method used).