Investment Themes: Core Value, Hedge
Imagine you have built a stock portfolio over several decades. You already rely on the dividend income provided by the portfolio. Then, the coronavirus happened. Because of the economic disruption your income declines and you will have to find new ways to cover your expenses.
What if there was a way to prevent this from happening? A hedge is an asset or security that performs well when the rest of your portfolio does not. Flow Traders ($FLOW.AS) is one of the stocks that skyrocketed due to the despair on the financial markets in March and April (not because it sold products online).
Today, I want to see if a liquidity providers like Flow Traders is still fairly valued to fulfill this role in a (100% equity) portfolio. There are only a few companies in this business listed on public exchanges, so it is hard to get exposure to the sector. I plan to analyze a peer in this market (Virtu Financial ($VIRT)) at a later point in time.
Flow Traders (I will refer to them as Flow in the remainder of this post) was founded as recently as 2004 and can be characterized as a financial technology (FinTech) company. Flow is specialized in Exchange Traded Products (ETPs), of which most of you probably know the Exchange Traded Funds (ETFs). However, there are many other types of ETPs traded on exchanges around the world. For example, whereas ETFs are often used to track stock indices, Exchange Traded Notes (ETNs) often track a commodity.
So what does Flow do, exactly? Flow is a market maker (MM) and authorized participant (AP) in ETPs. This basically means they quote bid and ask prices both on and off the stock exchanges to make sure people can constantly trade the ETP securities as well as the underlying securities that make up these products. In their own words, Flow is a liquidity provider and their business leads to greater efficiency and more transparency on the financial markets.
Their aim is to make money on each transaction. Their business model can therefore be broken down in two parts: trading volume and the bid-ask spread. The more volume is traded, the more money Flow makes. The higher the bid-ask spread, the more money Flow makes. This is a rather simplified representation of their business model, but it catches the main principle. Because their trading strategies require a lot of intelligence from both a human capital and technological standpoint, their technology platform, the trading strategies themselves and employees are Flow’s greatest assets.
With its origins and large market share in Europe, Flow was able to expand into the Americas and Asian markets, where it is a relatively small player. Yet, its global reach and technology enabled operations can still offer growth opportunities for the future.
High Frequency Trader (HFT)
Flow also falls within the category of High Frequency Traders (HFTs). These companies are often criticized by the public. Since these companies operate in complex ecosystems and in a sort of ‘black box’ on the financial markets, the debate on their role in, and impact on, the financial markets is still ongoing.
Firstly, HFTs can access market data sooner than other market participants. Some people see this as arbitrage and consequently consider their business practice unfair or unethical (especially because they ‘capture’ the bid-ask spreads from regular investors). Secondly, the trading algorithms are thought to create the infamous ‘flash crashes’, where these algorithms suddenly sell-off large amounts of assets. Lastly, the vast increase in investments through ETPs have become of public interest as well. Some researchers think the practice of ETPs and the market structure is unstable, especially when a lot of investors suddenly want to sell their assets (Michael Burry from ‘The Big Short’ even mentioned this was similar to the CDO markets). The Market Makers and Authorized Participants should then still be able to provide liquidity.
BlackRock, one of the larger ETP Sponsors in the world, posted an article supporting the resilience of the ETP market. The Federal Reserve Bank of Boston wrote an article in 2018 that seems to support BlackRock’s overall statements. However, there are some specific market developments mentioned in their article that seem to be overlooked in BlackRock’s piece, such as index-inclusion effects on individual stocks and co-movement of stocks, the increased concentration of the asset management industry into passive funds, and volatility amplifying investing strategies through leveraged and inversed ETFs.
Because of the large amounts of volumes traded and the low trading margins, risk management is at the center of Flow’s business. Flow repeatedly emphasizes the importance of risk management in their reporting and holds more than the regulatory required amount of capital at the end of almost every quarter (‘excess capital’). In the first quarter of 2020, when the largest volatility hit the financial markets due to the coronavirus pandemic, the company stated there were no ‘loss days’ (days with a negative trading margin). Since their IPO in 2015, Flow has posted a profit in every single quarter, further proving the resilience and profitability of the business model.
Flow covers different products and asset classes. In the 2019 annual report it stated coverage of all asset classes (equity, fixed income, commodities and FX) in ETPs and futures. Flow has some coverage, and is expanding its coverage, in spot products and FWD/NDF (forwards and non-deliverable forwards) product categories.
Many but not all ETP Sponsors and exchanges list their Authorized Participants and/or Market Makers. But the ones that do, almost always have Flow on there:
- Japan Exchange Group
- Hong Kong Exchange
- London Stock Exchange Group
- New York Stock Exchange
- Borse Frankfurt
Other companies that are also mentioned on almost all these lists are major players in the markets and are peer specialized firms such as Optiver, IMC and Virtu Financial. The FIA EPTA (European Principal Trading Association) provides a more comprehensive overview of the companies that are active in this industry.
Focus on Shareholders
Something I like about Flow is that the co-founders of the company, Roger Hodenius and Jan van Kuijk, are still a part of the Supervisory Board AND still hold a large share in the company. Javak Investments holds 12.22% of the outstanding shares, and Avalon Holding holds 10.07% of the outstanding shares (based on the 2019 annual report). To me, this shows that the founders still believe in the future potential of the company.
Furthermore, the company has a solid dividend policy to return at least 50% of the net profits in dividends to shareholders and it encourages employees to take a stake in the company’s equity as well. Lastly, Flow started a share repurchasing program this year on top of the existing dividend policy. As per the 31st of December, 2019 the outstanding share count was 46.5 million. As of the Q3 earnings release, this number decreased to 45.1 million (about a 3% decrease).
Investors in Flow seem not to be rewarded since the stock became a public company in 2015. The stock itself showed a lot of ups and downs, with a low point below €20 in 2017 and near the end of 2019. However, at some points, when volatility hit the markets, the stock showed some recovery to levels above €30 (most notably after the first quarter of 2018 and more recently the first quarter of 2020).
After the earnings for the third quarter, investors seemed to throw their Flow stocks back into the trash can again, with the stock tumbling from around €32 back to around €28. I hope to find out to see what the actual value of the stock is and to try to find some guidance through the heavy price movements.
Flow is active in the ETP market, which is a market that has shown a lot of growth since the start of this century. Bloomberg reported in September 2019 that the assets in index funds and ETFs surpassed those in active funds. According to CNBC, the passively managed funds grew from around $1 Trillion in 2010 to just over $3 Trillion by the end of 2018, while actively managed funds only grew from around $2.5 Trillion to $3.5 Trillion over the same time period. The market share between actively and passively managed funds further shows that passive funds still have a growth path ahead of them.
Flow posts an update with market statistics every month. In this update they estimate the total value traded in ETPs (both on and off exchanges) across different continents (EMEA, Americas and APAC). The graph below shows the growth that they are seeing in the market, starting in 2016. It is important to note that turbulence on the market has a large impact on the total value traded. In terms of volatility, 2017 and 2019 were very calm and 2018 and (especially) 2020 were more volatile. Also note that, due to the size of the market in the Americas, I used a logarithmic scale.
The ETP Process
Now let’s dive into the way the market works, to get a little bit more feeling of how Flow actually makes money. Remember that not even all academics know exactly how the market is impacted by all of these trading activities, so fully understanding this is probably either impossible or a fool’s errand (to use Elon Musk’s words on LiDAR). That is why I will rely on the flow charts included in BlackRock’s article, and a simple example of how a Market Maker and/or Authorized Participant can make money through this system.
After reading this, please let me know if you had expected that trading on the stock market was actually this complex behind the scenes!
The Process – Secondary Markets and High Frequency Trading
Let’s start with the simple basic process on the stock market, which BlackRock calls the secondary market trading activity. The image below is the image from BlackRock’s article, but I have added the grey ‘HFT’ blocks, standing for the High Frequency Traders that intervene between buyers, sellers and the exchanges.
Imagine you want to buy a stock for $35. Then on the other side there is a seller who wants to sell the same stock for $34.90. You would expect the exchange to match either the buy price of $35 to the selling price, or the selling price of $34.90 to the buying price. However, HFTs have nanoseconds of earlier access to the prices on an exchange, and can thus intervene in this process. They can buy the share from the seller for $34.90 and sell them to you for $35 to match the supply and demand, netting the $0.10 difference.
HFTs can do the same thing with shares trading on different exchanges and/or in different currencies. For example, if I want to buy a share on the New York Stock Exchange for $35 (or €30.10) that someone else wants to sell on a European stock exchange (where it is also listed) for €30, these traders can use this arbitrage in the market to net the difference.
With a technological advantage this may sound easy to achieve, but eventually the increase in technology expenses for the HFTs will drive their margins down substantially. Therefore, their competitive advantage needs to come from a combination of different trading strategies. This is where, at least for Flow, the primary market of ETFs comes into play.
The Process – ETF Share Creation and Redemption
Contrary to the image shown above, the image below is a one-on-one copy of the flow charts from the BlackRock article on the creation and redemption of ETF Shares. This is the regular process for ETF Share creation and ETF Share redemption, which can vary among regions.
It is important to understand that the number of ETF Shares on the market is flexible. Whereas a company usually has a fixed set of shares on the market determining its market capitalization, the number of shares on the market for an ETF can vary on a daily basis.
The ETF Sponsor is usually the one you know from the name of the ETF, such as iShares from BlackRock, Vanguard, or the SPDR ETFs from State Street. These ETF Sponsors use APs, Authorized Participants, to bring the ETF Shares on the market. The AP gives them the underlying securities in return for the ETF Shares, which the AP can then sell directly on the market to either Market Makers or investors. The other way around, ETF Shares can be redeemed by the AP when there is less demand in exchange for the cash or securities from the ETF Sponsor.
A Market Maker does not necessarily have to step-in, but usually provides – in the words of Vanguard – a ‘hidden’ layer of liquidity to the market by buying and selling ETF shares when there are not enough buyers or sellers on the market. This also optimizes the matching between the price of the ETF and the value of the underlying securities. Authorized Participants can be, but do not have to be, Market Makers and vice versa.
I would not be surprised if you would ask: ‘what’s in it for the Authorized Participant and/or Market Maker’? Well, let’s look at that now with another simple example.
We have an index that is tracking three stocks, all with a value of $10. At the start of the day, all stocks are trading at $10 and the ETF is trading at $30. The underlying value matches the value of the ETF Shares. Now one stock increases 10%, one stock drops 20%, and one stock stays unchanged. This means that the underlying value of the ETF drops from $30 to $29. Now let’s assume the index as a whole decreases by 2.5%. This means the ETF is priced too high. The Market Maker or AP could buy the underlying basket for $29 and sell (short) the ETF shares, netting the $0.25 difference.
As mentioned before, Flow is an Authorized Participant and a Market Maker, and is characterized as a High Frequency Trader by many (which is as far as I know not a real objectively identifiable label as almost any professional trading company these days would be a HFT to some extent). These market positions are key to their trading strategies. The examples above are very simple, and the practice is probably vastly more complex and slightly different (as is the case with many textbook examples).
Market Dynamics (Porter’s Five Forces)
The average number of APs per ETF ranges from 31 for small ETFs to 38 for large ETFs. This is a very small number of market players, especially considering the increased concentration in the passive fund market compared to the active fund market. This, combined with the trading complexity, capital requirements in the financial sector, technological edge and the necessary relationship (trust) with one of the large ETF Sponsors and/or Stock Exchanges, leads me to believe there is a strong entry barrier to this market.
Buyer power is very weak on the stock market, especially with a strong increase of smaller individual investors (that only make less use of active managers with more power). The smaller amount of suppliers (the ETF Sponsors) may be a threat, but more and more ETFs come on the market, making the role of a good AP and Market Maker more and more important.
The substitutes for investing in ETPs are currently not attractive, and are not likely to become attractive in the future either. Active funds have much higher fees, and investing in individual stocks is not something many investors want to spend their precious time on. Also, with the expectation for low interest rates the attractiveness of savings accounts also massively decreased over the past decade with many European banks even starting to charge negative interest rates above certain amounts of deposits.
Something I want to dive in a little deeper is the competition between incumbents in the market for APs and Market Makers. The competition between these companies is heavy, especially when market circumstances are not positive (low bid-ask spreads and relatively low volumes). In 2017 and 2019, earnings of Flow were also remarkably low, allegedly because of low volatility (measured by the VIX). The VIX is something I want to dive in a little later. First, I want to look at consolidation in the industry.
According to the article from BlackRock, 62% of their Primary Market Activity is divided over five different APs, with the top 3 holding a 44% share. The remaining 38% is held by 37 other APs. This implies the market for APs is very concentrated. Unfortunately, I could not find data to see the development of this market concentration over time. However, the data provided by BlackRock on Europe shows the concentration is even larger there, with 72% of their Primary Market Activity being divided over five different APs, with the top 3 holding a 60% share. With consolidation already happening in the industry (with the acquisition of KCG Holdings, Inc. by Virtu Financial as an example), the strong power of the larger APs can be expected to remain.
Flow reports their ETP and non-ETP value traded on a quarterly basis for the different regions. With their estimation of total ETP value traded, an approximation of their market share can be derived. For EMEA and the Americas they report the numbers quite clearly. For APAC, they changed the reporting in the second quarter of 2018 (from APAC to APAC excluding China). As they separately report the estimations for the Chinese market, their market share in APAC excluding China can be derived starting in Q2 of 2018. The data for EMEA and the Americas goes back to the start of 2016.
This data shows that Flow is one of the market leaders in Europe. The cause of the drop in their market share in 2018 is unknown to me, but the market share seems to be stable and increasing over the past few quarters. In APAC and the Americas, Flow is a smaller player. However, even though the market share is small in the Americas, the share of the Americas in Flow’s Net Trading Income (NTI) is much larger due to the massive size of the ETF market in the Americas (as shown above).
The theory on many forums and in many articles is that with high volatility ($VIX) and high uncertainty, the bid-ask spreads increase, which in turn offers more profit opportunities for APs and Market Makers. However, the VIX is nothing more than a metric for the expected (!) volatility in the stock market over the next 30 days. This means that, if uncertainty (expected volatility) is high but constant, the market may in reality not move that much and actual volatility may be lower.
This is something that, in my opinion, became very clear in Flow’s Q3 earnings report. The VIX was 26.05 on average during the third quarter of 2020. In comparison, the VIX was on average between 15 and 16 during 2019, with 16.55 in the first quarter of 2019. However, Net Trading Income came in at €78.4 Million, only slightly higher than the €63.1 Million of the first quarter of 2019 (which thus had a much lower VIX). To emphasize the shock-effect of this result, with an average VIX of 30.49 in the first quarter of 2020 Flow reported a Net Trading Income of €494.9 Million.
Therefore, what I believe is a much better metric to look at, is the change (!) in expected volatility. The more sudden the change, the more people will act on them. Just to make this clearer, let’s refer back to the same quarters mentioned in the previous section. The first quarter of 2019 saw a difference between the lowest point and the highest point of the VIX of 13.08 compared to the third quarter of 2020 with 12.02, which is quite similar. However, the first quarter of 2020 saw a difference between the highest and lowest point of the VIX of 65.50!
Note I got into all of this before getting into the financials and the valuation, because the financials are very volatile due to the nature of the business. It is very important to understand the business first with such a company, because recent financial performance can, in my opinion, be distorting in the view of Flow’s fair valuation.
Normally, you would value a stock based on a forecast of their earnings growth, cash flows and a discount rate. Now look at the image below and tell me what the earnings growth and future cash flows are going to be for Flow next year.
I think valuing Flow like a regular stock is impossible. Hence, this time I will not look at balance sheets or cash flow statements (due to the nature of Flow’s business their net profits are often very close to their cash flows, making them quite a good proxy for valuing the business).
The way I would like to approach Flow is with a dividend discount model (in perpetuity) based on the expected average long-term earnings at this point in time and their dividend pay-out ratio policy of a minimum of 50%. The only things I would need for this are the long-term Net Trading Income, the resulting Net Profit and the Enterprise Value.
As you would know by now, I would like to leave the discount rate as the unknown factor as it is the most subjective metric in my honest opinion.
Net Trading Income
The Net Trading Income for Flow is available starting from the first quarter of 2014. However, how can we estimate a long-term sustainable NTI for the company given the data we have at our disposal? To do this, I decided to put my assumption (the relation between the movements in the VIX and the NTI) to the test.
To forecast a long-term NTI I decided to build two regression models. Both models have the NTI as the y-variable. The first model uses the standard deviation in the daily VIX data (that is available through Yahoo Finance) as the x-variable. The second model adds a second x-variable, being the average VIX (close) over that time period. The two models showed some deviation to both the upside and the downside. As with many things in life, the truth often lies in the middle, so I took the average outcome of the two models as the definitive outcome. This led to the following result of output (green) versus actual NTI (red).
Of course there are still outliers to the upside, and outliers to the downside. However, I am personally quite satisfied with the results, especially for the outliers in NTI and the differences during ‘average’ quarters.
Now I can use this model historically, using VIX data from Yahoo Finance going back to 1990, to see what Flow would have done historically based on this model. The outcome was rather surprising. Not a single quarter would have been as profitable as the first quarter of 2020. The third quarter of 2008 would come in second with €271.19 Million, compared to the €491.47 Million of the first quarter of 2020. The average NTI per quarter based on the model would be €73.94 Million, and the median would be €59.87 Million.
Something interesting to note in the graph above, showing the quarterly output of the model, is that the number of outliers on the upside seems to increase over time. The average over the 30-Year period (starting Q1 1990) increases from €73.94 Million to €79.72 Million and €82.62 Million for the 20-Year (starting Q1 2000) and 10-Year periods (starting Q1 2010), respectively. The median remains quite unchanged with €62.08 Million for the 20-Year period and €60.54 Million for the 10-Year period. As I like to be conservative, I will continue to use the 30-Year figures (also, more data means more information).
Net Profit and Dividends
Flow uses a lot of variable employee benefits to manage their profit margins and to incentivize the employees to continue to get the best trading results. As a result, the net profit margin varies a lot, but is still relatively stable. The average net profit margin for Flow from the first quarter of 2015 up until the third quarter of 2020 was 33.6%, and the median net profit margin over the same period was 32.4%. Based on both the median and average outcomes of the NTI and Net Profit Margin, this leads to quarterly Net Profits between €19.40 Million and €24.86 Million.
Based on the number of shares outstanding of 45,103,408 at the end of the third quarter of 2020 and a dividend pay-out ratio of 50%, this would lead to dividends between €0.86 and €1.10 per share on an annual basis. The earnings per share in those cases would be €1.72 and €2.20, respectively.
Enterprise Value and Implied Returns
The market capitalization is currently €1,268.31 Million (or €1.3 Billion) at a share price of €28.12. Flow reported they had excess capital of €276 Million at the end of the third quarter of 2020. Let’s assume, for the sake of simplicity, this is equal to their net cash position (as a financial company). In that case, the Enterprise Value of Flow would be €992.31 Million. The Enterprise Value per share would then be exactly €22.
Given a pay-out ratio of 50%, this would imply a dividend discount rate between 3.9% and 5.0% based on the Enterprise Value per share. In case the pay-out ratio would increase to 100% – which would then be equal to the total earnings yield – this would naturally double to a range between 7.8% and 10.0%.
Some things to consider when looking at this implied return, is that with current market circumstances it is still likely that Flow will earn more than this long-term amount over the upcoming years. Volatility has historically led to more volatility, and the low interest rates in current market circumstances make the market much more sensitive to new information. With the VIX touching the 40 levels again over the past week, this implies Flow may have a Q4 that is much better than their Q3 figures. The time value of money teaches us that this might make Flow more valuable at this moment, increasing the implied return today.
Another remark is that the 50% dividend pay-out ratio is below the historical average, and the no-growth assumption furthermore assumes that the retained 50% does not return any more additional earnings compared to the current base case scenario. Flow has been using its capital to invest in new initiatives, such as a stake in the Members Exchange. Additionally, it is still making moves to increase its market share in the United States and the (still growing) Asian markets. Given these initiatives and the outlook for a dividend yield between 3.9% and 5.0% on the long-term (average of 4.5%), this makes Flow seem quite an attractive stock (in my opinion – not advice) after the drop caused by the disappointing Q3 earnings, especially given the current market circumstances.
Given the uncertainty in the financial markets at this moment, the current valuation and the remarks that need to be taken into account when valuing Flow Traders, I am still very happy to have them as my largest holding in my individual stock portfolio.
What do you think? Have I changed your mind on buying a negative beta stock in the depths of the ETP markets? Did I overlook anything that would further support the investment thesis, or did I miss something that would disprove it? Let me know in the comments!
P.S.: watch my Twitter feed going forward, as I will try to predict Flow its quarterly earnings using the model I built and the VIX data.
P.S.2: don’t get scared of the volatility in the markets: happy Halloween!