We write a lot about return dispersion with respect to stocks – the phenomenon whereby stocks either all move together, or begin to act distinctly. Dispersion comes and goes; when it disappears, like it did eight years ago, passive investing makes a great deal of sense. The return of dispersion in 2017 has been a boon to active managers, who are benefiting from it and other similar factors allowing for positive spreads vs. the market. No surprise then that through Q1 2017 more than half of all active mutual fund managers have beaten their 2017 benchmark.

Dispersion is the mother’s milk of active management, and skilled active managers can significantly enhance returns during such periods, highlighting a shortfall of passive investing and so-called “smart-beta” ETF’s. Our partners at FlowPoint Capital create research models that highlight dispersion opportunities at the market, sector and stock level. One of their methodologies we track closely is FlowPoint’s Trend1 model, a trend-following system that categorizes stocks into four groups: A, B, C and D. The graph below shows the performance of Russell 2500 stocks in these categories since 1990. These data highlight (1) amplitude and duration of the four categories’ share price performance over time, and (2) the power of a systematic approach to portfolio construction.

The Problems With ETF'S, Part II- Junk in the Trunk

Source: FlowPoint Capital

But for a brief “worst-to-first” run such as in 2008, stocks in the C and D categories are serial capital destroyers. Bull markets, bear markets, risk-on, risk-off, no matter – C’s and D’s are to be avoided. Well, guess how many stocks in the S&P 500 are in the C and D categories today? Eighty-six, or almost 25%. To be fair, many C-rated stocks have positive returns YTD, but the number of D’s with positive YTD returns can be counted on one hand. This in a year when the market is up 10%. So why own them? Well, you do. In your ETF. If you are holding your ETF and its C’s and D’s because you’re waiting for 2008 again, we should talk.

Today, D’s include a heavy representation of Consumer stocks and to a lesser extent, Energy. The carnage in the Consumer space has been well documented, with more than 35% of that sector negative for the past year, and Consumer is a 21% weight in the S&P 500. Nearly 100% of the Energy holdings in the S&P 500 are down YTD, though 100% of FlowPoint’s A-ranked energy stocks are positive for the year. More than 20% of FlowPoint’s category D stocks are down at least 30% YTD, more than a 40% return dispersion from the market’s. You can do the math many different ways, but what your return could have been by eliminating known risks like C’s and D’s is material.

That’s the opportunity cost with passive vehicles. You have to take the D’s with the A’s – like buying a car with a motion activated lift gate but having to also pay for the sun roof you don’t want. This risk is presumably what “smart-beta” ETF’s reduce. But smart-beta is a misnomer. Tilting portfolios toward lower-beta stocks, or high dividend payers, is one way to reduce portfolio volatility or enhance yield, but that simplistic approach doesn’t go far enough.

Take a hard look at your ETF’s. Then, build, buy or rent ranking systems that work and give your ETF an x-ray. See what’s inside. Your portfolio will thank you for it.

Crow Point would like to thank its intern, T.J. Pavone, for his help in compiling this post. It was TJ who did most of the research and analysis here. He has redeemed, somewhat, the millennial generation with his hard work. May he be an example to others.


In the past few weeks, we have teased a few thoughts on the issues we see in ETF’s.  This week, we’re going to start a series called “The Problems in Your ETF.”  Today’s post centers again on the levered ETF’s.

Many (negative) market commentators point to Central Bank purchases of stocks as a looming issue – i.e., Central Banks have no business buying stocks, their purchases have helped push stocks to artificially high levels, and when those purchases inevitably reverse, stocks will get crushed, etc, etc.  Given their status as a buyer of last resort, their tendency so far to be long term holders, we’re not sure why Central Banks are bad shareholders.  But that’s beside the point, which is: why does no one talk about the skew introduced into the markets by the levered ETF’s?

Once again, our friends at FlowPoint Capital have supplied us with the necessary research. FlowPoint looked at 788 ETFs in the US that have greater than $1.0 million in average daily trading volume.  The total average daily value traded in this group is $65.6 billion, or 28% of all US equity trading.  The total market cap of these ETFs is $2.9 trillion, or only 9.4% of the entire US equity market capitalization.  You starting to see the disconnect?

The average ETF in this universe trades four percent of its market cap per day, and the average holding period is 162 days.  The $235 billion SPY, for example, trades eight percent of its market cap every day, which amounts to a 15-day average holding period.  In contrast, its buy-and-hold Vanguard cousin, the $80.0 billion VTI, has a 340-day average holding period.

The problem really is with the geared or levered ETF’s.  While some non-levered ETF’s have holding periods of 700 days, $93.0 billion in ETF market cap averages a holding period less than ten days, and most of that is in levered ETF’s which are designed specifically to encourage daily trading.  Does this concentration in levered ETF’s presage a potentially looming problem for equity markets?   Here it is graphically:

Source: FlowPoint Capital and Bloomberg

Look at the geared Direxion gold miners, the DUST (3x short) and NUGT (3x long).  The DUST has $273 million in market cap but trades $189 million per day.  The NUGT has $1.5 billion in market cap but trades $300 million per day.  Its swaps are based on the Vaneck gold miners’ ETF (GDX).  But, not too long ago, stories in the press appeared noting the issues GDX was having meeting subscriptions and redemptions. And that was just because the GDX was growing rapidly.  What happens when volatility spikes and everyone runs for the door?  What about the swaps the NUGT has on the GDX?  You know why no one knows?  Because neither the NUGT or the DUST existed in 2008.  They were created in 2010 and have only known stable markets.

The TQQQ, your favorite 3x levered technology play, owns $1.8 billion in tech stocks, including $100.0 million in AAPL.  It was also only formed in 2010.  When all the fish start swimming in the other direction, what will that leverage unwind do to stocks?  Our point is that the number of shares owned today by short term shareholders is ominous.  Levered Bull funds outweigh levered Bear funds in number and in assets. So, while it’s nice to have the Bear funds to supply liquidity in Bear markets, they will get swamped by their cousins. And it isn’t just the levered vehicles.  Some of the widely owned country and sector ETF’s have shockingly high turnover rates, which will only get worse in a prolonged selloff.

Remember when the Flash Crash in ETF’s was an issue?  That was two days in August 2015.  What happens when it turns into something real, when people are actually paying attention?  It’s easy to pick on the levered ETF’s.  The real concern ought to be with the un-levered, widely held ETF’s.  Many names in the ETF universe, mostly created after 2010 and untested in bear markets, attract tourists and not investors.  That’s a problem. Portfolio insurance, a brand-new product and untested in Bear markets, exacerbated the 1987 crash. Long-Term Capital, everyone’s favorite pariah, was a brand-new concept when it almost caused markets to melt down in the late 90’s. Risk parity funds are brand new. And so are many ETF’s.

Last week, amid some controversy, the SEC approved a 4x levered ETF. For a period of time, and rightly, the SEC was considering cracking down on the 2x and 3x levered vehicles so the move to 4x was surprising. Don’t misunderstand, the levered ETF’s offer gigantic arbitrage opportunities because of their structural problems, so in many respects we hope they never go away (we are invested in a fund that benefits greatly and almost daily from these arbitrage opportunities), but that doesn’t mean they are great products for the audience for which they were created.

Bloomberg recently reported that passive ETF indexes now outnumber individual stocks. Without getting into the multitude of problems created by the explosion in passive vehicles, the one overlooked point seems to be how many of these passive ETF’s, because of supply and structural issues, must use the futures markets instead of underlying securities to create their exposure, and therein lies the problem. Remember, a huge issue leading up to and exacerbating the 2008 crash, was the growth in usage of Credit Default Swaps (CDS). There are many parallels to today’s ETF markets.

For those of you unaware, CDS initially were born as a way for commercial banks to hedge their lending exposure. CDS were quickly latched onto as a way to solve a chronic problem in high yield: a lack of paper. If you were a high yield portfolio manager pre-2000, and you had a great idea about how undervalued a credit was and you wanted to buy a bond, often you couldn’t because there were simply no bonds available to purchase. It was not uncommon in the heyday of high yield for a $300 million new issue to be 80% owned by three holders with scraps available for the rest of the market. Enter CDS.

If one could create synthetic exposure, a high yield portfolio manager could reflect almost any view of credits and credit markets, arguably a good thing. And, like many good things, CDS quickly morphed into Frankenstein, and synthetic securities grew to be multiples of the notional value of the underlying securities and that leverage partly fueled the 2008 unwind.

The same situation exists with many ETF’s today, especially the commodity ETF’s, whose problems are exacerbated by contango. When a fund purchases forward to meet the creation requirements of the underlying ETF, and forward curves are more expensive (contango), value leaks away. It is a mathematical fact exemplified by the chart below, which shows that Crude Oil (the blue line) is up marginally in 2017 while the USO (the yellow line), the ETF that represents the price of oil, is down. It’s down because of the leakage in value it suffers when it sells a future for a lower price than the future it must purchase. For a period in 2016, Crude was up 27%, but the USO was down 1%. You can thank Contango for that result. And the USO isn’t even levered.

The levered funds magnify the contango effect geometrically. Capturing the gap between the value of the underlying ETF and the commodity it represents is commonly called “roll yield” and billions of dollars are in the market today arbitraging exactly that. As long as contango exists, this situation will persist. Sooner or later, something will give. We thought the regulators would step in and quash the levered ETF’s. For now, they’ve chosen not to. In the meantime, the market is giving you an opportunity.

In practice, it works like this: Assume a 1% daily move to make the math easy. Day one the market goes down 1%, the next day up 1% and so on for ten days (see table below). Under that scenario, the underlying security returns -0.05%, while both 2x levered vehicles would return -0.20%, the 4x levered vehicles -0.80%. Shorting both the long 4x and the short 4x would return 1.59%. The contango decay as outlined above adds a large measure of downside protection. Two of three potential outcomes are positive: if the underlying sells off, or if it trades with volatility. It is no wonder these strategies are becoming more popular. The odds are in your favor. Adding a long position to protect against an upside spike lowers the risk profile of this trade even further.

Source: Flow Point Capital and Archimedes Fund Management