Description

Long: Virtu Financial (VIRT)

Share price: $28.00; Target Price: $63.47; ~130% upside

Introduction

VIRT is a financial trading firm. Market making firms like VIRT are the reason that brokerages are able to support commission-free trading. VIRT has been an unloved stock recently, but trades at extremely attractive valuations with multiple potential catalysts creating potential for 130% upside. VIRT has a 5.3B mkt cap on a 6.95B enterprise value and trades at $45M/daily.

Brokers like Robinhood route their orders through firms like VIRT, who match all the buys and sells while keeping a small spread, with retail traders getting better prices commission-free. Giving some of this spread back to the originating brokerage is known as Payment for Order Flow (PFOF).  Beyond market-making (60-80% of NTI), VIRT also earns revenue by providing execution services (20-40% of NTI) to financial clients like banks. Basically, the higher the market volume and volatility, the more VIRT earns in its market-making segment, with execution services generally staying stable throughout.  

Last year’s volatility was great for VIRT, with earnings exploding upwards to $6.38 per share (LTM from 1Q21), compared to $1.15 in 2019. But with the stock trading at $28, or 4.7x ttm PE, investors seem to be pricing in significant decreases in earnings for the foreseeable future, primarily based around a return to stock market normalcy or regulatory headwinds.

I disagree. I think that COVID has masked significant and permanent changes in the retail trading landscape and VIRT’s own business which are currently unappreciated by the market. On a normalized basis, VIRT trades at an 8-15% FCF yield, assuming no further improvements to the industry, nor additional organic growth initiatives. Over the long-term, we think that there is sufficient scope for further expansion.  Furthermore, there is a good chance that the regulatory risk is overblown, considering the strength of retail incentives to protect VIRT’s business model. Over the next few years, management’s focus on shareholder returns via share repurchases should help drive this overlooked and underappreciated stock upwards. To proceed, I examine the bear case in detail and show why these arguments are incorrect.  I then build out a more likely picture of VIRT’s path forward.

The bear case

The consensus bear case has three prongs:

1. 2020 was a one-time thing. The market is rapidly recovering, and reopening will mean hard comps, lots of stimulus spending, and general optimism. Volume and volatility will dissipate, and VIRT’s earnings will return to a lower level. Current prices reflect a cyclical peak. 

2. VIRT was unprofitable in 2019, and will return to being so once conditions “normalize”, especially because market-making is a commodified business with bad economics.

3. PFOF is a politically contentious issue and will be subject to regulatory intervention, especially given new SEC Chair Gary Gensler’s recent remarks, where he suggested that there was a trade-off between PFOF and improved customer prices. 

Response

1. VIRT actually stands to benefit from substantial secular tailwinds that are masked by COVID

2020 was indeed a great year for volatility, but 2021 is holding up well. 2021Q1 was actually VIRT’s strongest quarter to date, with quarterly EPS of $1.89, and NTI of ~$12m/day despite significantly lower volatility than all quarters of 2020. The key driver: volumes. To be more precise, retail volumes have exploded, with retail brokerages like Interactive Brokers reporting more than 4x the average daily volume compared early 2019, and VIRT handling ¼ of all retail investor orders in the US.  While some of the volatility from COVID and the craziness of GameStop, AMC, etc., is undoubtedly temporary, I suspect that a significant amount of the increased volume from the past year or two is here to stay. 

Firstly, the adoption of commission-free trading by discount brokerages in 2019 was already driving Daily Average Revenue Trades (DARTs) and users upwards well before March of last year. Retail investors are hugely dependent on commission-free trading as a result of their small account sizes(Robinhood’s average account size is $3,500 for instance), and combined with the secular trend of online platforms like Reddit, TikTok, and Twitter actively forming social communities/cults around stock ownership, I think that this retail involvement has legs.

Secondly, Millennial’s and Gen-Z also trade far more frequently than their older peers, with about a quarter of millennials saying that they day-trade stocks, almost 4x that of their boomer parents. Over the long-term, a good proportion of millennials will be receiving a massive amount of wealth from their Boomer parents and grandparents over the next decade or two. Based on current indications, a good proportion of that money will be invested far more actively than before. VIRT, who currently commands 30% of retail volumes in the US, is undoubtedly poised to benefit from this, and also increased ownership of certain asset classes like crypto.

Finally, a large section of global assets, especially in credit, has yet to transition to a mostly electronic system of trading. As firms like MarketAxess work on increasing liquidity, VIRT will see increased demand for its execution and market-making services. Additional opportunities also exist on the international front, with many markets like the EU and Japan being relatively untapped despite the substantial size of their economies. The combined potential of both could more than double VIRT’s addressable market.  

2. The market-making business is misunderstood

The most common perception about trading-related businesses is that they are vulnerable to huge losses, and have no moat because they are simply making risky bets, or are at risk from someone stealing their algorithms.

The market-making industry is different. Firms like VIRT and Jane Street have always made trading income, and often significant operating profits as well. The majority of them have rarely suffered huge losses, and only from idiosyncratic, tail-risk errors. One would think that with ROICs>20%, firms would be entering in droves, but there have essentially not been any new market-makers since the financial crisis; the majority, like IMC, Jane Street, Optiver, and Citadel were founded in the 1990s. These market makers often have highly-siloed access to code, built up massive repositories of data over two decades, and integrate specialised, self-improving programs with proprietary, customised equipment/chips, forming real and tangible barriers to entry. With barriers to entry this high, it would not surprise me if the market continues to consolidate around 1- 2 players(For instance, VIRTU and Citadel have 22.8% of the total US stock trading market share, more than the NYSE, Nasdaq or Oboe)

3. 2020 disguises VIRT’s substantial cost-cutting measures and organic growth since 2018/2019

Even if the overall environment returns to their pre-commission-free, pre-COVID, pre-Meme stock days, VIRT has made significant improvements on to their cost structure. VIRT’s 2019 reported numbers were muddied by the acquisition of solutions provider ITG to reduce its reliance on volatility and volume. Since 2019, the combined entity has cut over 100m in fixed costs, even when employee compensation was severely larger in 1H2020(as a result of the extreme volatility). Management is guiding towards 600-650m in indirect/fixed expenses post 2019, which works out to another ~200m below FY2020, or 2H2020 annualized (implying that most of the cutting has probably been achieved), which means that the combined entity will have 2/3rds the fixed cost as compared to 2019. The most recent quarter was at $180m, and that was including substantial excess compensation from COVID etc. Over the long-term, they have guided to $600m in fixed costs. In addition, management has achieved over $166m of “organic growth” in NTI since 2018, or approximately $500k a day. Management has stated that 

they think that they can hit ~$1m of incremental NTI per day than 2018 or 700k more per day than 2019. 

Estimating VIRT’s steady - state earnings

Explanations of assumptions are in the footnotes, but in short, VIRT can probably produce levered FCF of between $1.8 to $3.7 per share in a bear case and a bull case respectively, with the base case at $2.4/share. 

Even if we assume the absolute worse-case scenario of low volatility [e.g., 2019], and ignore the additional effects of organic growth or the secular effect of commission-free income, VIRT should be able to earn almost ~$2 of FCF/share [post a bit of deleveraging/share repurchases]. If we add in the additional $700k of NTI, we get NTI of $4.7m/ day, or ~$2.5FCF/share, almost 10% yield. If we consider 3Q2020 as our “new normal”, taking $5.7m/day of NTI + 300k of incremental growth initiatives, we have a 15% FCF yield, or $3.84/share of EPS.

Needless to say, I think the base case is far more likely. As noted in the footnotes below, the 2019/bear case takes a trough year and haircuts the previously achieved “organic” growth by more than a third. The doubling of retail investor accounts with commission-free trading combined with the current volatility levels mean that we are unlikely to see previously trough year levels of NTI any time soon. The base case is more reasonable, as it assumes some degree of elevated volatility/volumes on par with the lowest of the most recent quarters, and is more likely to stay here for the next few months if not years considering current retail involvement and interest.

Naturally, the current market is anything but normal, and VIRT just earned ~$2 in EPS in this past quarter. Should this elevated volume persist, VIRT could earn almost 20% of its market cap in FCF this year.

An Undemanding Valuation

VIRT has no publicly listed competitors, making comps rather difficult. Its seemingly closest comps are probably firms like CBOE and MKTX (22x and 66x P/E respectively), all of which operate different businesses from VIRT. Brokerages IBKR and SCHW have done well in a banner year, but also trade at highly elevated valuations as well (46x and 32x P/E respectively). Even with VIRT’s higher debt load and lower growth, VIRT still appears undervalued if we consider the “New Normal” case, as the firm pursues growth initiatives and de-levers.

What makes VIRT’s valuation even more interesting is the potential for repurchases. Outside of paying a 3.5% dividend yield, management has indicated that they are comfortable with the debt level and would prefer to repurchase shares over anything else. With founder Vincent Viola holding 36% of the stock and sovereign wealth fund Temasek holding 3% of shares, the float is smaller than it first appears, making these steady purchases even more impactful. Additionally, with 76m shares being held by the top 10 shareholders, trading dynamics could result in an even larger impact on the share price. Management has indicated that after accounting for the $0.96/share or $182 dividend [<1/2 of Q1 profits], most of the excess cash will be used to repurchase shares. 

On an absolute level, the base case valuation of ~8.8 P/LFCF seems particularly low given the quality of the business and the shrinking float, with the bear case current multiple of 11x P/LFCF still being significantly below its peers. 15x P/FCF on the “new normal” case gives us a target price of $63.47 or ~130% upside. TEV/EBITDA values are also low, ranging from 12.3 to 7.2 TEV/EBITDA. Naturally, all bets are off if this period of exuberance continues, with VIRT earning >2$ a share of FCF[>$400m] per quarter as of today.

We can also create a conservative sensitivity table for our FCF/share and TEV/EBITDA. Even the worst-case scenario looks pretty alright-applying a 12x P/FCF multiple gives a base price of 19.32, or a 30% downside. Note also that NTI levels since 2020 have stayed well above the NTI values in this table.

Risks

Regulation

HFT will always be controversial, and this is not VIRT’s first time dealing with negative publicity. Remember the controversy surrounding Michael Lewis’ Flash Boys book? Or the Swedish transaction tax in 1994? Fact of the matter is, attempts to inhibit the HFT/market making space have never worked out. 

Even after recent events surrounding the GameStop mania, most serious proposals seem to focus on regulating Robinhood’s gamification. There is a recently proposed bill on the transaction tax, but support seems to be tepid for now, especially considering the lack of political willpower and the administration’s more centrist roots, to say nothing of the potentially disastrous consequences for the financial industry. Gensler’s remarks add some negativity to the potential tail risk, but I hardly think that the effect is that significant.

There is a possible risk that PFOF could be banned, but that would simply result in firms like Robinhood having to charge a commission on a per-trade basis - something which retail investors would loath and could be somewhat politically tricky – remember the whole outcry over GME/AMC? There is also a tangential risk that brokerages like Robinhood build their own HFT firm, but that seems rather difficult if not filled with conflict of interest. VIRT CEO Doug Cifu mentioned that it would “take years and a nine-figure investment”, to say nothing of the inherent structural barriers as mentioned earlier. Even if PFOF were banned, I would not be surprised if some of the secular trends around increased stock involvement and awareness are here to stay, to say nothing of the potential for electronification.

Balance sheet/business model

There is another risk to VIRT via potential trading losses, which are always tricky to anticipate given the black box nature of these algorithms. Historically, the main reason why market makers have lost money is due to software/human error. For instance, Knight Capital (which was bought over by VIRT), lost $400m after a software engineer failed to copy new code. (See this link on the Knight Capital disaster for an example of how to not have risk management controls in software implementation)

While there is no good way to know exactly whether this will happen, I think that the chances are fairly unlikely. Knight Capital, as the SEC and others found, had very little in the way of safety controls and risk management tools. Furthermore, the error is not so much a feature of the business as it is a human error-perhaps no different from how a bank investor would consider “fat-fingered” payments as a risk, or how we have, “mine safety disclosures”. I view these events as unlikely edge cases as opposed to structural industry risks. I also find it interesting that the bonds (due in 2026) trade at par at a ~5% interest rate, so at least the credit investors don’t seem to be particularly worried about it, which is particularly interesting given the substantial losses that can happen if errors like these are allowed to spread.

Event Path

The way I see it, the current meme stock rally either continues grinding out, or suddenly collapses. But regardless of whichever happens, VIRT wins with either volumes or volatility. Perhaps more importantly, VIRT’s earnings post-market collapse is likely to normalise at a higher level than previously, which could lead to some degree of multiple expansion as investors come to appreciate the increased resiliency and secular trends ahead. Should it keep continuing, then VIRT will continue chugging along, returning significant amounts of cash-if VIRT earn somewhere in between Q12020 and Q12021 NTI levels for the rest of the year, management has indicated that they will repurchase ~15-16m shares at current prices, ~11-12% of the float.

Conclusion

The value of VIRT over the next 12 months is not in the just the relative undervaluation, the overlooked and difficult to understand business with lowered costs, or even the fact that the stock has achieved >20% ROIC in the past few years and operates in a market which has had basically no new entrants in the past decade while its main business risk is idiosyncratic and generally uncorrelated with the broader market.

Should this era of euphoria continue, then VIRT will continue earning outsized returns and return capital to shareholders. And when this period inevitably ends, we will find ourselves with an 8-15% steady-state FCF yield for a negative-correlation, long-volatility asset which will benefit from secular trends ahead. What’s not to love?

Catalyst

Positive Catalysts

1. Q2 earnings - VIRT has beat earnings in the last four quarters, and may do so again at current volumes

2. More buybacks - Management has said that they will repurchase more shares if profits hold steady 

3. COVID coming back/any other major shock to the market/blow-ups

Risks

1. Transaction tax/regulatory surprises/ PFOF being banned

2. Blowups-i.e. Knight Capital

3. A low-volatility, steady and “boring” stock market- alternatively, should VIRT continue to earn decent profits as in the base case and “new normal” above, multiple re-rating could be dramatic and swift