|Shares Out. (in M):||177||P/E||78.6||61.65|
|Market Cap (in $M):||12,201||P/FCF||27.8||25.6|
|Net Debt (in $M):||-226||EBIT||202||261|
|Borrow Cost:||Available 0-15% cost|
Fool’s Gold: Why Franco-Nevada is a Short
Introduction and Thesis:
As the price of gold has risen over the last year Franco-Nevada (TSX: FNV, “FNV” or the “Company”) has experienced a significant appreciation in its stock price, rising from a low of $41.47 to a high of $81.16, and recently trading around $70. As a result of this advance, barring a material additional increase in the price of gold, Franco-Nevada is dramatically overvalued relative to its mineral reserves and expected future cash flows. Additionally, with little product differentiation and increased competition as a result of past success, future growth for the business should be much harder to come by. If you anticipate further inflation in the price of gold, you are better off with a more fairly valued product with similar commodity exposure. If not, FNV is an attractive short.
By way of corporate background, the Company succinctly summarizes its recent history in its most recent 40-F.
Many of FNV’s assets were originally acquired and developed by Franco-Nevada Mining Corporation Limited (“Old Franco-Nevada”), Normandy Mining Limited (“Normandy”) and Newmont Mining Corporation (“Newmont”). Old Franco-Nevada was a publicly-listed company on the Toronto Stock Exchange (the “TSX”) from 1983 to 2002 and had originally acquired royalties at the Goldstrike complex along with many other royalties. In February 2002, Newmont acquired Old Franco-Nevada along with Normandy. Old Franco-Nevada’s assets and part of its management team were incorporated into a new division of Newmont called Newmont Capital Limited (“Newmont Capital”). Newmont Capital’s activities included the management of royalty, investment and project portfolios as well as corporate development. Pursuant to an acquisition agreement, FNV acquired its initial royalty portfolio from Newmont effective December 20, 2007 and also agreed to assume certain liabilities related to such portfolio. FNV then continued to build its asset portfolio through additional acquisitions, adding numerous royalty, stream and other assets since 2007.
FNV is a leading precious metal royalty and streaming company. It offers financing to miners, paying them upfront for a claim on a portion of the production of a mine. The Company does not engage in any mining activity itself, but is instead focused on managing its pool of royalty and streaming assets. Royalties entitle FNV to a share of the minerals mined or revenue yielded from a mine with no additional payment required from FNV. Streams also grant FNV a claim on a portion of the resources generated, but require payment from FNV, usually at rates that are dramatically below current market prices, to purchase the asset. To a limited degree the Company also purchases Net Profit Interests (NPIs) entitling FNV to a share of the profits from a particular mine. Selling royalty and streaming rights can be attractive for miners seeking a middle ground between debt and equity. Royalties and streams provide upfront cash but do not establish preferred creditors or scheduled payments like debt, and setting aside the agreed upon reduction in cash flows, are not dilutive to the shareholders like equity. FNV measures its mineral reserves by converting expected future cash flows into Royalty Equivalent Units (“REUs”), which measures the value of a cash stream as if it were coming from a gold royalty. For example, a stream expected to generate 100,000 ounces of gold for FNV, but requiring payment of 50% of current market value, would be worth 50,000 REUs. Using this metric allows for comparison across different types of metals and contracts. The Company divides REUs into categories based on the certainty of the existence of the resources and the economic viability of bringing them to market given current prices. Proven and Probable (P&P) reserves probably exist, and are realizable at current market prices. Measured and Indicated (M&I) reserves also likely exist (encompassing P&P reserves), but also include reserves that are not feasible to develop given current commodity prices. Inferred reserves are less likely to exist and may not be economically viable to bring to market.
Although FNV was one of the original pioneers in this industry, competition has intensified in recent years. The largest direct public competitors are Silver Wheaton (TSX: SLW, “SLW”) with a market capitalization similar to FNV, and Royal Gold (NasdaqGS: RGLD, “RGLD”) with a market capitalization significantly below the two market leaders. FNV is also in competition with other potential capital suppliers, namely the debt and equity markets. Given these rivals are competing to provide financing for future streams of cash, there is little to differentiate them. In spite of their limited competitive advantages, FNV and its competitors have seen significant price appreciation of their common stock over the last 10 years, with FNV and SLW achieving a roughly 300% increase and RGLD advancing by roughly 200%.
FNV bulls are largely banking on three factors to justify current valuation and drive future price appreciation. First, given the upfront nature of payments and limited additional capital requirements, FNV is a cash flow machine on existing royalties and streams. In 2015 FNV generated $339.3 million in adjusted EBITDA on just $443.6 million of revenue, an incredible 76% margin. This allowed the Company to increase its dividend to a total of $129.0 million, or more than 25% of revenue. Taken at face value, these cash flow dynamics are outstanding, and one reason investors have driven the stock higher in recent months.
Second, FNV exhibits attractive optionality features, with limited downside (and almost no risk of total loss) given its low operating costs and limited debt, and seemingly significant upside given its exposure to a broad mix of undeveloped properties with future development possibilities. FNV maintains little or no debt on its balance sheet, and generally pays either nothing or a rate significantly below the market price to acquire gold from its royalty and streaming deals. Barring substantial contingent liabilities of an unforeseen nature, there is almost no chance that the Company goes broke in the near future. Additionally, with royalty and streaming claims on undeveloped land in major mineral development regions, the Company is exposed to potential discoveries on those lands. For these reasons management consistently pitches the Company as one with limited downside and meaningful prospects for price appreciation as new discoveries organically develop.
Finally, FNV’s management has a history of creating value for shareholders as exhibited by the Company’s stock price appreciation and growth over time. From 2011 to 2015 Gold Equivalent Ounces (GEOs), or the total publicly reported Mineral Reserves and Mineral Resources on properties in which Franco-Nevada has interests, have grown from 237,722 to 360,070, or over 50%. Since 2008 FNV’s revenue has nearly tripled as the Company has added new streams and royalties. As stated above, over the last ten years the stock price has roughly quadrupled. Given managements long history in the industry, and these attractive results in the recent past, the market appears to hold the current team in high regards and expect similar results going forward.
Diving into the meat of the bear case, it is straightforward based on the future cash flows the Company expects to generate, and FNV’s current reserve position, that it is materially overvalued.
Digging down to a mine by mine analysis of the Company’s future gold and silver production over the next 40 years paints a bleak picture for buyers of the equity at the current valuation. Assuming a 10% discount rate (see argument below) and a flat gold price (in real terms), the discounted value of all cash flows over the next 40 years totals just $3,349 million. Combining that with $226 million of cash on the balance sheet and dividing by the current share count yields a total value per share of $20.11, or over 70% below the current share price! In total the Company is expected to generate $13.5 billion in revenue with $2.7 billion in COGS, $0.6 billion in SG&A (excluding depletion), $0.9 billion in taxes, and $0.7 billion in capital expenditures. The remaining factor to arrive at $3.4 billion reflects the discounted nature of the future cash flows. REUs produced total approximately 7.6 million ounces, similar to the P&P reserves accounted for by the Company. However, the model uses the assumptions for production at the Cobre Panama mine as laid out below, meaning that setting Cobre aside, the assumptions in the model actually more closely track the more aggressive M&I reserve provided by the Company. Being generous and assuming a 5% discount rate and $1,475 average price of gold, FNV still only has an intrinsic worth of $33.70 a share, less than half the current price. Most shockingly, at current gold prices, even assuming a 0% discount rate, intrinsic value is only $49.56 a share or nearly 30% below the current stock price!
Examining total mineral reserves does not improve the story for FNV. Proven and probable REUs total 7,639,000 ounces. Multiplied by the current price of gold and added to cash, this tallies to approximately $10.2 billion compared to a current market capitalization of $12.2 billion. Obviously focusing on reserves alone ignores the role of corporate expenses (approximately $20 million annually over the last several years) and taxes (26.1% of profits at the statutory rate per the most recent annual report). Combined, those expenses are expected to total over $1.5 billion over the next 40 years, equivalent to more than 1.1 million REUs. Additionally, focusing on reserves alone ignores the timing of cash flow and the discount factor that must be applied to estimate intrinsic value. Finally, the totals given so far include questionable REUs from Cobre Panama (discussed below). Removing those reserves from the equation leaves just 5.7 million ounces of P&P REUs, worth just $7.4 billion. After removing operating expenses and taxes, and discounting to the present, the bear case for FNV becomes apparent.
In response to the above statistics the Company might respond that a significant portion of their anticipated future revenues is not captured by REUs, and that the optionality present in their existing holdings will create more REUs over time. Upon closer examination of the recent past, this claim appears dubious. After accounting for annual production and excluding the impact of new acquisitions on total REUs, P&P reserves have shrunk three of the last four years, and in total have shrunk by over 2.4 million ounces, or about 32% of current reserves. Over the same time period, excluding the impact of new mines and annual production, M&I reserves coupled with inferred reserves have shrunk all four years, and by a whopping 8,323,000 ounces in total. That totals an incredible 61% of current M&I and inferred REUs worth almost $11.0 billion. Moreover, these adjustments in REUs are spread across the entire base of assets and don’t appear to reflect just one or two projects going away. Even after including acquired reserves the Company still has seen a decrease in REUs per share as a result of the significant number of shares added to pay for acquisitions. At year-end 2012, with the price of gold at $1,675 an ounce, FNV had 48.87 P&P reserve ounces, 68.48 M&I reserve ounces, and 32.01 inferred reserve ounces per thousand shares. By the end 2015, with the price of gold down to $1,060 an ounce, those totals had decreased to 42.96, 61.69, and 14.54 respectively. Yet incredibly over the same time period the stock price had increased from $56.78 to $63.30, and now stands at around $70. Instead of enjoying the benefits of optionality on its holdings, FNV instead appears to be bleeding value each year as its existing assets are marked down and its stock is diluted!
Finally, even setting aside the inability of the Company to generate new reserves from its existing pool of assets, the characterization of FNV as an option on the price of gold seems overstated. Given most costs, including COGs, are fixed for a given level of production, a change in the price of gold, either positive or negative, flows directly to the bottom line. Therefore, the Company actually offers exposure to gold fairly similar to that found in the commodity itself. FNV is no silver bullet. Exposure to mining upside comes with substantial threat of loss as a result of commodity price risk.
Problems with Current Market View on Business Model:
Issues with the current valuation are not limited to the mathematical calculations above. Although there are elements of truth to each of the bulls’ arguments, there are significant problems with portions of the long view on the stock, and important other issues with the Company’s business profile that the market appears to be ignoring.
First, investors focus on free cash flow is completely misguided in the case of a mining company with depleting assets. Insofar as the Company expended capital to acquire the royalties and streams that it derives free cash flow from, it is most appropriate to include depreciation and depletion in any calculation of profits used to derive a valuation. Claiming that FNV generates a “3% free cash flow yield” based on EBITDA/Enterprise value is foolish when 2.5% of that must be set aside for new streams and royalties to maintain that cash flow. As a comparative demonstration of this divergence for FNV, it is useful to compare it’s EV/EBITDA and P/E ratios. With an EV/EBITDA ratio of ~30x bulls point to a 3% free cash flow yield. However, with a P/E ratio of ~115x FNV appears to be generating a yield around 0.8% after accounting for taxes and the replacement cost of depleting royalties.
Shockingly, even this metric for free cash flow materially overstates the value being created for shareholders at the current price. Looking toward Benjamin Graham’s recommendation in Security Analysis for valuing mining companies (FNV would qualify considering its economic value is derived from mines), he says, “Where the life of a property is limited, the stated depreciation charge should be ignored and the “investor’s amortization” charged against the earnings before depreciation” (Security Analysis, 2nd edition, Graham and Dodd, pg. 496). For Graham, “investor’s amortization” is calculated simply as the current enterprise value divided by the minimum or probable life of the mine. The idea is that if an asset is losing 5% of its initial value annually, the cost to investors is 5% of their cost basis in the asset, not 5% of the original cost of the asset. Modeling out the cash flows from each individual mine in FNV’s portfolio yields an average time of the cash flows of 13 years for the portfolio as a whole. Doubling this metric assumes average probable mine life of the portfolio of 26 years. Dividing current enterprise value by 26 yields an “investor amortization” charge of $457.3 million compared to an LTM depletion charge of $249.2 million. Subtracting the $457.3 million “investor amortization” charge from adjusted earnings before depreciation and depletion yields a pre-tax loss of $68.2 million. At the current price investors actually lost 0.6% of their capital last year on an economic basis! Quite a significant difference from the already paltry 3% free cash flow yield advertised by bulls. Standing alone, free cash flow as defined by EBITDA or cash from operations is irrelevant to FNVs economic value and should be ignored as a red herring.
Second, FNV has almost no competitive advantage compared to other potential providers of capital. In a bidding war with SLW and RGLD, the winner will be the firm willing to provide the best terms. Put another way, the firm willing to accept the lowest return on its invested capital. While at one time this space may have been underserved, providing attractive opportunities for capital suppliers, it now appears to be overserved (for examples, look at the liquidity positions of FNV and its competitors, and at the bond yields of major gold miners). Quoting directly from FNV’s 2015 annual report,
Many companies are engaged in the acquisition of mining and oil & natural gas interests, including large, established companies with substantial financial resources, operational capabilities and long earnings records. Franco-Nevada may be at a competitive disadvantage in acquiring those interests, whether by way of royalty, stream or other form of investment, as competitors may have greater financial resources and technical staffs… Franco-Nevada may be unable to acquire royalties or streams at acceptable valuations which may result in a material and adverse effect on Franco-Nevada’s profitability, results of operations and financial condition.
FNV’s CEO admitted that finding new deals in the space is problematic right now with the equity and debt markets wide open to miners in an interview at the Mines and Money Conference on September 28th. Given the hunt for yield around the world, increased competition in an industry built on financing capital projects seems assured. There is no reason to forecast an abatement of this intense competition in the near future. Although this does not spell certain doom for FNV’s business model, it does imply that maintaining a 3.5x Price/Tangible Book may be unrealistic when the other major suppliers of capital, namely banks, trade closer to 1.0x.
Third, it is important to consider the incentives of mining companies working with FNV to finance capital projects. Miners always have the option of issuing debt or equity to raise capital as opposed to selling royalties and streams. It seems unlikely that miners would be willing to give up streams with an IRR of 10-15% when they can issue debt at a fraction of that price. For example, Barrick Gold Corporation (TSX: ABX) has long term bonds outstanding with current yields under 5%. Even with a weighted average cost of capital (“WACC”) of 4.5%, as assumed by Deutsche Bank in a recent research report, FNV will have a hard time competing with alternative providers of capital while maintaining its 3.5x Price/Tangible Book Value ratio. Additionally, although the FNV team employs a group of experts to evaluate new opportunities, it seems unlikely that they would know more about the projected return from a given royalty or stream than the company selling them said royalty. The notion that FNV is getting cut rate deals on properties with massive reserves and discovery opportunities appears to be based more on hope than fact or logic.
As an example of the above two factors at play, one need look no further than at the recent Antamina rights acquisition. In 2015 FNV paid $610 million for a stream with total P&P REUs of just 349,000 ounces worth just over $462 million. With mine life expected to extend to just 2028 per the Company’s most recent annual report, and annual production expected to total 2.8-3.2 million ounces of silver attributable to FNV over that time period, a simple calculation using the 3.2 million ounces estimate derives a production estimate of about 466,000 REUs, worth about $750 million. In an upside scenario FNV generated an IRR of just a couple percent on one of their largest acquisitions in the recent past! Coupled with low implied IRRs at other recent acquisitions, FNV appears to be materially lowering its standards for deals, just as the arguments above would predict.
Finally, in assuming a 4.5% WACC to justify FNV’s current valuation, analysts are not incorporating the significant exposure to developing markets captured in FNV’s portfolio of royalties and streams. Just 26% of FNV’s M&I REUs reside in the United States, Canada, or Australia as opposed to 74% in developing countries within Latin America, Africa, and the Middle East. Some of these countries have poor track records of respecting property rights for western investors. Although there does not appear to be short term risk to any of the assets, it is unusual for a company so dependent on risky mining ventures in remote parts of the developing world to have such a low cost of capital. Even a cursory glance at historical precedents would indicate that the odds of some kind of political risk developing on a portion of FNV’s assets in the next 25 years is high.
Additionally, fully 33% of FNV’s measured and indicated resources come from just one mine, Cobre Panama, which is still under construction in Panama. There are always significant risks associated with any major construction project, particularly one as large as Cobre Panama. In this case, there is reason to believe that not only are there significant operational risks associated with the project, but also that FNV has been extremely aggressive in crediting itself with REUs from the project. Per the latest annual report from First Quantum Minerals (TSX: FM), Cobre Panama has an expected life of 30 years. Per FNV’s latest annual report, over those 30 years it anticipates generating 328,000 tons of copper annually for the first 20 and 228,000 tons thereafter. FNV in describing its arrangement states,
Gold deliveries are indexed to copper in concentrate produced from the project. 120 ounces of gold per every 1 million pounds of copper produced until 808,000 ounces of gold delivered. Thereafter 81 ounces of gold per 1 million pounds of copper produced to 1,716,188 ounces of gold delivered, thereafter 63.4% of the gold in concentrate. Silver deliveries are also indexed to copper in concentrate produced from the project. 1,376 ounces of silver per every 1 million pounds of copper produced until 9,842,000 ounces of silver delivered. Thereafter 1,776 ounces of silver per 1 million pounds of copper produced to 29,731,000 ounces of silver delivered, thereafter 62.1% of the silver.
FNV will pay $406 per ounce of gold and $6.09 per ounce of silver. Based on the above statistics, FNV is expected to receive approximately 1,872,600 ounces of gold and 32,562,560 ounces of silver over the next 30 years. After netting the cost of these ounces and translating to REUs this totals just 1,478,187 ounces. This compares to a P&P reserve of 3,403,000 ounces and a M&I reserve of 3,657,000 ounces of REUs as stated in the Company’s latest Asset Handbook. Over 2 million ounces of REUs are unaccounted for and impossible to back into based on available public information! Best case it appears there are significant undisclosed factors at play here that will determine the viability of approximately 20% of FNVs total reserves. Worst case, there is a significant and undisclosed hole in the balance sheet of this Company. Either way, a 10% cost of capital would seem much more reasonable (if not still overly aggressive) for a company with significant commodity exposure, political risk, operating risk, and murky financial support for its most important asset.
Risks to Short Position:
Although the above (hopefully) presents a fairly compelling argument for a short position on FNV, there are several risks to consider in establishing a short position. First, and most obviously given the exposure to gold, is a dramatic increase in the price of gold. It would take a significant increase in gold prices to justify the current valuation, but a large run up in gold would likely lead to short term losses and potentially a long-term loss if sustained. If you believe that a large rise in the price of gold is likely I would suggest either avoiding this short or using it as a hedge with a fairly valued gold-exposed instrument to capture the upside while limiting the downside of the position.
Second, it is possible that the Company could raise enough new equity at or near the current valuation that the stock’s intrinsic value rises to that level. Using a simple and exaggerated example, if the company issues 100 billion shares at its current value for cash, the old delta in price vs. intrinsic value would be swamped by the overwhelming amount of new equity, effectively closing the value gap for a short seller. Historically the Company has raised most of the cash it has used for new acquisitions from shareholders, making this a real potential risk if attractive assets came to market in the near future. However, it would take a series of major acquisitions over a number of years for this threat to really deal a blow to the thesis, so assuming the market realizes its mistake on this stock over the next year or two an attentive short seller should be alright. Also, in a downside case, issuing new equity at the current stock price doesn’t justify a higher price for the stock, meaning that a short theoretically shouldn’t lose money.
Finally, it is possible that the Company could find large reserves on its existing properties, dramatically increasing its intrinsic value over time. Although this is a possibility, the Company has a poor track record of increasing the value of its current holdings over the last four years, as discussed above. Even if a group of large mines were discovered, given the current valuation is arguable over 3x intrinsic value, the set of cash flows from the new discoveries would need to dwarf existing holdings for this to be a real long term threat to the thesis, particularly given the average time to realize the value of those cash flows would likely be longer than from the existing pool of assets. This seems highly implausible, and given FNV’s focus on growing through acquisitions of new royalties and streams in the recent past, not something that the Company appears to be counting on going forward.
FNV is not a stock that is going to zero. It is not overleveraged, it generates substantial cash flows, and it appears to have decent and competent management. However, as the case above lays out, it is dramatically overvalued at present, based both on expected future cash flows, the value of mineral reserves, and the increasingly competitive nature of the business. Although any company trading at 100x earnings can usually be classified as a speculation, given the depleting nature of FNV’s assets and the low competitive advantages present in its business model, the market appears to be expecting something far beyond perfection from the Company. In summary, with a total annual cost to short of less than 2% annually, a short with a price target of $30 a share (still 50% above estimated intrinsic value from cash flows) seems appropriate and conservative. Like FNV’s business model, such a bet seems to have very limited downside and potentially significant upside.