|Shares Out. (in M):||45||P/E||0||0|
|Market Cap (in $M):||315||P/FCF||0||0|
|Net Debt (in $M):||-127||EBIT||0||0|
|TEV (in $M):||193||TEV/EBIT||0||0|
Sign up for free guest access to view investment idea with a 45 days delay.
GAIN Capital Holdings (NYSE: GCAP)
GAIN Capital Holdings provides trading technology and execution services to retail and institutional investors globally across multiple asset classes, including foreign exchange, commodities, equities, options, and futures. GAIN’s business includes a retail FX/CFD business spanning eight regulatory jurisdictions, a U.S.-based retail futures business, and an institutional trading business (GTX).
lpartners wrote up GCAP in Sept 2015, and I have added numerous comments to the message string there, but given developments since then and that the stock is still extremely cheap on both an absolute basis and a relative basis, I thought it would be worth adding additional context about the opportunity.
There are some valid questions about the quality of the retail FX/CFD business (discussed below), but I believe concerns about those issues have led to a broadly negative opinion that has driven all optimism out of the price and that the consensus psychology incorporated in the GCAP price is too negative. The retail FX/CFD business model has some issues given the level of customer churn and need to replace churned customers via marketing or acquisition in order to grow. However, that does not mean that GCAP is a “sell” regardless of price.
“At some price, every company is a buy; at some price, every company is a hold; and at a still higher price, every company is a sell.” - Seth Klarman
“There is no asset so good that it can’t become overpriced and thus risky, and few so bad that there’s no price at which they’re a buy (and safe).” – Howard Marks
I believe customer churn in this business is perceived to be even worse than it actually is (discussed in detail below). GAIN could do a better job of reporting detailed metrics about customer economics (e.g., customer acquisition costs, ARPU, customer lifetime value, etc.), which is partly a driver of this misperception, but its competitors in the UK (IG Group, Plus500, CMC Markets) do report detailed customer metrics (included below), which give a better understanding of the customer economics in the business.
The investor concerns about the retail FX/CFD business have led to a discounted valuation for the entire company, which has obscured the value of several hidden sources of value for which the issues with the retail FX/CFD business do not apply:
1) GTX: the institutional business (worth an estimated >$145 million)
2) The Retail Futures Segment (worth an estimated $35-40 million), and
3) The company’s net cash balance, which is misunderstood due to the manual adjustments needed to calculate it ($127.4 million at 12/31/17—$285.7m cash and equivalents, and $158.35m debt).
If you assign GTX a 15% discount to the multiples of the BATS / Hotspot acquisition (equivalent to a 25% discount to the multiples of the Euronext / FastMatch acquisition), that implies a conservative estimate for the value of GTX is >$145 million. With a conservative valuation for the Futures business in combination with GAIN’s excess cash, GAIN’s retail FX/CFD business is almost trading for free:
GAIN’s retail FX/CFD business did $27 million of EBITDA (even with allocation of all corporate overhead to the segment) in 2017 (a year that was penalized by volatility near multi-decade lows leading to the lowest LTM EBITDA in years), and $90 million of EBITDA net of all corporate overhead in 2016, a year that was not strong in historical context (also low volatility). A valuation of $12 million for that business is far too cheap.
Even if you believe the retail business is in terminal decline, and you model decline curves to calculate the NPV for the retail business, the discounted value of the future free cash flows is worth something meaningfully more than its implied value today. GAIN’s unlevered free cash flow in 2016 (a year with fundamentals not strong in historical context, but stronger than 2017) was $45 million: a 23.4% free cash flow yield based on today’s EV.
I believe GCAP at the current price has a large margin of safety: it does not require that things to improve in the retail FX/CFD business for an investor to do well; it is so low that you can at least avoid loss, and can potentially profit, even if things get worse in the retail FX/CFD business.
Based only on the results of 2016 and the low end of GAIN’s public comps, the retail business is worth close to $450 million, which along with the values of the other segments, implies a value for the stock of over $16.50 (vs. $6.99 now). In addition, there are several drivers that should improve GAIN’s results and profitability over time (e.g., increased currency volatility, rising interest rates, increasing interest rate differentials across currencies, which will drive increased carry trade: see the “Catalyst” section for more detail).
Finally, GCAP includes a free option on any upside from adding bitcoin and other cryptocurrency trading (which, to be clear, is not part of my thesis about GAIN’s undervaluation, but cryptocurrency trading has driven improved results for GAIN competitors). In Dec 2017, GAIN announced the launch of Bitcoin trading to City Index U.K. clients (either as a spread bet or as a CFD), and it has since expanded the offering to Ethereum, Litecoin, Ripple, and Bitcoin Cash in the U.K., Europe, Australia and Singapore. They plan a U.S. launch this year. In the US (where CFDs and spread betting are not allowed), GAIN’s Futures division will be offering bitcoin futures, which are strictly cash settled, with no actual ownership of bitcoin. IG Group and Plus500 have both announced that their results have been helped by strong volumes in CFDs on cryptocurrencies. CMC Markets launched CFDs and spread betting for cryptocurrencies in March '18.
I’ll first highlight the sources of value obscured by the concerns with the quality of GAIN’s retail FX/CFD business, and then will discuss the quality concerns about the business.
Key Sources of Obscured Value
1) Cash Balance
GAIN’s cash balance and cash flow statement are a bit confusing and misunderstood due to movements in the “Receivables from brokers” line item, which are funds posted with brokers as collateral for holding trading positions. They represent GAIN’s cash, simply on deposit with brokers.
Both GAIN’s cash balance and “Net cash provided by operating activities” need to be adjusted for the “Receivables from brokers” and “Payables to brokers” lines, with “Receivables from brokers” added to the cash balance, “Payables to brokers” subtracted from cash, and with operating cash flow adjusted for changes in those accounts—sometimes increased, and sometimes decreased (e.g., an increase in “Receivables from brokers” on the Balance Sheet will cause a corresponding decrease in reported operating cash flow, which should be added back to operating cash flow because the increase in “Receivables from brokers” represents a real increase in GAIN’s cash and cash equivalents).
One source of the confusion is that services like Thomson Financial, Capital IQ, and FactSet do not correctly report GAIN’s net cash because of the manual adjustment needed to account for the “Receivables from brokers.” They only give GAIN credit for the “Cash and cash equivalents” line, which now understates GAIN’s cash by $76 million. If sell side analysts give GAIN any credit at all for their cash balance in comparable tables, they also almost always only give them credit for only the “Cash and cash equivalents” line.
GAIN added to their quarterly reports a calculation of “Net operating cash,” and net of all debt, GAIN had $127 million of net free operating cash at 12/31:
December 31, 2017
Cash and cash equivalents $209,688
Cash and securities held for customers $978,828
Short term investments $305
Receivables from brokers (1) $78,503
Less: Payables to brokers ($2,789)
Net Operating Cash $285,707
Less: Debt Principal ($158,350)
Net Free Operating Cash $127,357
Excess Net Cash: $121,184 (net of 2% of Revenue)
(1) Funds posted with brokers as collateral required by agreements for holding trading positions.
Numerous sell side analysts that cover GAIN simply apply a P/E to their earnings estimates, and ignore GAIN’s cash balance completely, based on the view that their EPS estimates already assume productive use of that cash. But simply applying a P/E multiple to interest earnings from the cash gives little to no valuation credit for significantly different levels of cash, and drives a distortion when comparing companies with widely different cash balances. This issue is particularly accentuated with interest rates at their recent historic lows.
While there are regulatory requirements for the amount of capital GAIN must maintain, in an acquisition by a strategic acquirer, GAIN would get credit in its valuation for all of the $127 million of net cash, because a strategic acquirer would already have its own regulatory capital set aside, and would therefore be able to access all of Gain’s net operating cash.
2) GTX: GAIN’s Institutional Business
There has been ongoing consolidation in the institutional FX ECN business, with exchanges acquiring institutional FX platforms / Prime of Primes, and there are limited remaining independent players.
The complexity of institutional FX technology infrastructure helps explain the recent activity by exchanges pursuing FX acquisitions, which highlight the difficulty of entering the FX ECN market other than by acquisition. For example, NASDAQ wanted to launch NASDAQ FX but was unable to do so, likely due to the differences of liquidity provision in an OTC trading environment vs. an exchange-traded environment. NASDAQ FX did not have enough banks and FX market makers committed and connected to secure sufficient liquidity, and may have had challenges with the trading infrastructure needed to support the lifecycle of an FX trade. Another example: CME Group tried to push its exchange-listed FX products into the U.K. against the OTC FX firms in London, but pulled back from London and Europe after a year.
Euronext’s stated “Strategic Rationale” for acquiring FastMatch was the following: “The G20 post-credit crisis response has created a regulatory drive for Over-The-Counter (“OTC”) and dark trading towards transparent, neutral and centrally cleared markets. New capital requirements further underpin an urgent need for greater efficiency and deep structural change of the OTC landscape. Nowhere is this more relevant than in the $5.1 trillion daily foreign exchange market. This new environment, driven by regulatory changes and the client’s need for more transparency and efficiency has resulted in a clear trend of electronification of spot FX trading where 66% of trading in 2016 was electronic, growing from 55% in 2010.”
Acquisitions of FX ECNs have been the following:
Transaction data for these acquisitions:
Some recent articles discussing the acquisition interest from exchanges:
GTX Recent Financials:
GTX has been gaining market share in a market challenged by low volatility. GTX’s ADV during Q1’18 was $17.1 billion, slightly less than FastMatch’s ADV ($17.7 billion) when it was acquired for $181 million. GTX is larger in revenue than FastMatch was when it was acquired, but FastMatch was growing faster at the time. Hotspot was larger than GTX in revenue and ADV when BATS acquired it, and was growing at a rate similar to GTX in 2017. An acquirer could eliminate significant costs after an acquisition of the revenue stream.
If you assign GTX a 15% discount to Hotspot’s acquisition multiples (equivalent to a 25% discount to FastMatch acquisition multiples), that implies GTX is worth >$145 million.
3) Futures Segment
Futures Segment Recent Financials:
Futures Segment Comps:
Average Daily Contracts in GAIN’s Futures segment have increased in Q1’18 (Q1 average so far is 35% higher than Q4’17 and 45% higher than Q3’17). But even using a conservative multiple based on 2017 results (i.e., 11x 2017 EBITDA compared to the comps above at 11-16x 2018E EBITDA—excluding IBKR, which is not as comparable), a conservative valuation for the Futures segment is $35 million.
4) Retail Segment
Retail Segment Recent Financials:
Retail Segment Comps:
A key part of the bear case is that a high percent of FX and CFD traders lose money, which leads to customer churn, so there is a concern about the sustainability of the business. But IG Group, CMC Markets, Saxo Bank, and Plus500 are in the same business—so any concerns with GAIN’s business model should also apply to them—but they trade at far higher multiples than GAIN.
GAIN’s retail FX/CFD business did $27 million of EBITDA (even with allocation of all corporate overhead to the segment) in 2017 (a year that was penalized by volatility near multi-decade lows leading to the lowest LTM EBITDA in years), and $90 million of EBITDA net of all corporate overhead in 2016, a year that was not strong in historical context (also low volatility). Based only on the results of 2016 and the low end of GAIN’s public comps above (CMC Markets), the retail business is worth close to $450 million. In addition, there are several drivers that should improve GAIN’s results and profitability over time above the performance achieved in 2016 (see “Catalyst” section for more detail), which should drive the valuation beyond that level.
GAIN’s key competitive advantages are barriers to entry and their risk management capabilities.
Barriers to Entry in FX
There are barriers to entry that reduce the threat of new entrants from required technology infrastructure, government regulations / licenses required in each jurisdiction (both capital requirements and strict compliance requirements that are costly to implement and vary by country), economies of scale (there is a high degree of operating leverage in the business).
Technology Infrastructure that is Scalable Globally
Retail FX brokers need to build their own technology, which requires large investments of time and money but can result in competitive differentiation not available to retail equity brokers. The technology required for the lifecycle of an FX trade is different from that for equities. Because FX is traded OTC among large banks, FX market makers, and other financial institutions instead of on an exchange, it does not have a common trading infrastructure. So established retail FX brokers have built most, if not all, of their trade processing technology themselves. Standardized third-party infrastructure solutions are generally not available on a per trade basis, as in online equities. Although this increases the costs for a broker to enter and remain in the retail FX business, this differentiation also enables retail FX brokers to compete on the basis of the quality of their platform rather than merely on commissions per trade.
The complexity of the technology infrastructure helps explain the recent activity by exchanges pursuing institutional FX players, which highlight the difficulty of entering the FX market other than by acquisition, as discussed above.
In addition, retail FX is a global market, and the technology infrastructure is scalable globally, unlike that for equity trading. Trading of equities varies by country, often involving different lists of equities, different trading venues or exchanges, and different regulators. To trade equities globally, a brokerage firm generally must establish significant infrastructure in each major market. As a result, most online equity brokers source customers primarily from pools of investors within their own country. Unlike equities, spot FX contracts are neither traded on local exchanges nor cleared through a local clearing agent, and FX trading is generally similar around the world. So a retail FX brokerage firm does not need to build unique infrastructure in each market to offer trading services globally. In addition, instead of stocks unique to each country, the eight most popular currency pairs are traded by investors in many countries and represent over 72% of global FX trading volume. Because the retail FX market is essentially global, the potential market that is addressable by an online retail FX broker is larger than that addressable by an online provider of retail equities trading.
GAIN is subject to the regulatory requirements of regulatory bodies including the CFTC and NFA in the U.S., the FCA in the U.K., the FSA, METI, and MAFF in Japan, the SFC in Hong Kong, IIROC and the Ontario Securities Commission (OSC) in Canada, MAS in Singapore, ASIC in Australia, and the CIMA in the Cayman Islands, relating to liquidity and capital standards, and a variety of other requirements.
Regulators have made a series of changes that impact retail FX brokers, including substantial increases in minimum required regulatory capital, increased oversight of third-party referring brokers and regulations regarding the execution of trades. While these regulations may increase costs, another effect of these regulations has been to significantly reduce the number of firms offering retail FX, even as the number of customers and the volume traded has grown. As the industry has consolidated, scale has become increasingly important, presenting opportunities to larger firms that can meet the more stringent regulatory requirements.
Between 2010 and now, the retail OTC FX firms operating in America has been reduced from roughly 25 to three: GAIN, TD Ameritrade Futures & Forex, and Oanda (Interactive Brokers discontinued its US-based retail service Sept 1, 2016 for customers with <$10 million US as a result of a Dodd-Frank rule that only brokerages registered as Retail Foreign Exchange Dealers (RFEDs) with the NFA will be able to provide retail foreign exchange services).
The consolidation has been driven by a combination of regulation, acquisitions, and foreign firms pulling out of the market due to capital requirements and / or execution issues. The NFA capital requirements (which were also approved by the CFTC) for Forex Dealer Members (FDMs) increased to $10 million in October 2008, $15 million in January 2009, and $20 million in May 2009 (plus 5% of the amount (if any) by which liabilities to retail forex customers exceed $10 million). That requirement led numerous smaller and under-capitalized companies to seek merger partners with better capitalized companies, and both GAIN and FXCM have made numerous acquisitions (GAIN has completed ten significant acquisitions over the last five years).
Economies of Scale
More recently, after the Swiss National Bank’s decision led to negative client balances at numerous retail and commercial brokerages which process their order flow on an agency basis, counterparty credit has become a larger focus for liquidity providers, prime brokerage, and Tier 1 banks whose liquidity is aggregated by non-bank institutional providers which connect retail trading platforms.
As a result, it has become increasingly difficult for OTC FX brokerages to secure relationships with prime brokers for providing FX (and other electronically traded OTC asset classes) to retail customers as Tier 1 FX banks (Citigroup, Barclays, Deutsche Bank, HSBC, JP Morgan, and Goldman) have restricted counterparty credit to OTC firms. This has increased the capital requirements to maintain existing prime brokerage relationships and / or establish new ones. Just five years ago, it was possible to establish new and / or maintain existing prime brokerage relationships with Tier 1 banks with just $5 to $10 million in capital. Nowadays, a firm needs 10x those amounts on their balance sheets to establish new and / or maintain existing prime brokerage relationships.
GAIN’s scale offers also numerous other competitive advantages:
Risk Management Capabilities and Benefits of Principal Model
When the Swiss National Bank (“SNB”) discontinued its currency peg of 1.2 EUR per CHF in January 2015 leading to negative client balances and capital shortfalls at numerous brokerages (e.g., FXCM, Saxo, Alpari UK, etc.), GAIN actually made a profit that day.
Following the SNB decision, the EUR/CHF move was the largest move of a major currency since currencies started floating in 1971 (a change of over 40 standard deviations), but GAIN generated a profit for the day considering both trading profit and negative client equity, demonstrating that their conservative risk-management approach and the principal model position them to manage risk effectively during both calm and volatile markets. GAIN had increased client margin requirements for EUR/CHF to 5% in September 2014 in order to help mitigate negative client equity in the event of a major event.
The event also demonstrated the benefits of the principal model vs. the agency model. Following the SNB event, the failure of the banks to provide pricing resulted in some clients having negative balances, and when clients could not cover their margin call with their broker, brokers with the agency model still had to cover the same margin call with their liquidity providers and were liable to their liquidity providers for those shortfalls.
GAIN’s customer trades are either naturally hedged against an offsetting trade from another customer, hedged through an offsetting trade with one of their liquidity providers or may become part of their net exposure portfolio. For EUR/CHF customer trades in the net exposure portfolio that led to negative account balances, GAIN did not have associated margin calls with their liquidity providers.
Concerns with the Quality of GAIN’s Retail FX/CFD Business
Concern 1: Customer Churn
Customer churn is an issue in this business. The risks of highly leveraged trading in any asset class are large, and online FX brokers allow traders to take significant leverage with their positions, which can lead some investors to lose a large percent of their equity. Some traders trade FX until they lose enough money that they decide to stop trading FX, at which point may churn off the service and need to be replaced.
Organic demand is clearly preferable to a heavy reliance on marketing spend (all else equal), and businesses that have to acquire customers (either via marketing or acquisition) have a non-optimal business model that should and will trade at lower multiples than a company that does not have to heavily rely on marketing spend. As they grow, increasing marketing spend and/or acquisitions are needed to replace churned customers. If CAC rises (in this case it has not, and has actually fallen for some players: see the data from Plus500 below), total spending can increase to a point where shareholders lose patience with the amount of marketing spending necessary to fund growth. When marketing spending is reduced, if churned customers are not replaced, the company can begin to shrink.
Some have argued that the industry should not exist due to the level of customer losses. The other side of that debate is that individuals should have a role in self-directed trading in any asset class and should be able trade what they want to trade. Regulators can require reasonable disclosures and limits designed to protect investors, but the limits should not be to such a degree that they dramatically curtail public choice and kill an industry.
In addition, I believe customer churn in this business is perceived to be even worse than it actually is for at least three reasons:
a) CFTC Reporting requirements overstate the issue
Accounts that were flat (i.e., no loss and no profit) in a given quarter are included in the count of “unprofitable” accounts for the quarter.
CFTC rules require that the risk disclosure statement provided to every U.S. retail forex customer include disclosure of the number of non-discretionary accounts in the U.S. that were profitable and those that were not (including accounts that were flat) during the four most recent calendar quarters. GAIN’s current disclosure for U.S. accounts is the following:
Unprofitable (or Flat)
This account profitability is skewed downward by small accounts. Accounts can be opened for as little as $50, and small accounts are less profitable than average-sized accounts because these traders tend to take these accounts less seriously, with more aggressive with leverage and less conservative stop-loss protections. Larger accounts are more likely to be profitable, and have a longer tenure (see next section).
IG Group (one of GAIN’s competitors in Europe) reports more detail about client attrition than GAIN does, but GAIN’s customer data is likely similar. IG Group’s client attrition has been running at ~65% in year 1 and ~75% over two years, so monthly churn in year 2 was only ~0.9%: