Dr. Strangelove, Or: How I Learned To Stop Worrying And Love The Markets

|
Includes: CME, DHLSX, EDV, ICE, IEF, OTCRX, QLENX, TLH, TLT, TOTL, WHOSX
by: Kevin Wilson

Summary

The flash trading ecosystem has grown to dominate the markets over 15 years; the whole point is to gain an unfair advantage - so why is it tolerated?

A series of major market flash crashes in 2010, 2012, 2014, and 2015 indicates that regulators have inadvertently allowed systemic tail risk to permanently increase.

The notion that we can avoid situations where multiple forcing functions (investor psychology, overvaluation, algo trading, etc.) could combine to produce another 1987-style crash is epic scale wishful thinking.

Regulators have done nothing, but the new IEX Exchange (Private) at least stops front running technology and gives investors a fair shake; however, dark pools have greatly decreased market transparency.

Investors should avoid NDAQ, ICE, or CME; prudent investors may need higher bond (EDV, WHOSX, TLT, TLH, IEF, TOTL) and cash allocations, plus some long/short funds (DHLSX, OTCRX, QLENX).

Peter Sellers Movie Poster:

Source: dejaview.com

The Stanley Kubrick movie "Dr. Strangelove" was a 1964 satire about the Cold War and the concept of Mutually Assured Destruction ("MAD"), the defense strategy that purportedly kept everyone safe in a nuclear age. There were genuine fears back then that the alleged safety of "MAD" was a bit exaggerated. The movie was a much-needed attempt at humor; indeed I can remember being periodically ordered to shelter under my desk during Cold War air raid drills at my grade school in the late 1950s. The humor in the "Dr. Strangelove" movie involved insane generals and advisors, and complex systems that were "guaranteed" not to fail, but then did anyway, triggering the so-called "doomsday machine."

The film serves as a metaphor of sorts for the current situation in the markets, in my opinion. By that I mean that we have allowed uncontrolled algorithmic ("algo") traders to put us all at risk of a machine-generated market crash. We have been given all sorts of stories about how it's really good for liquidity, order flow, and trading costs. I will stipulate that all these things are true at some level, since there is decent documentation of it. But flash crashes happen with alarming frequency, and the markets appear to have amnesia about the granddaddy of all algo-related crashes, the Black Monday Crash of October 19, 1987. The crash was likely caused by a combination of market psychology, overvaluation, illiquidity, regional markets arbitrage trading, and program trading involving "portfolio insurance." While program trading was not the sole cause in 1987, it contributed, and the prudent reader may also notice that a somewhat similar suite of conditions may exist now.

Source: financial-spread-betting.com

On May 6, 2010 the world witnessed an epic "flash crash" that introduced many of us to the relatively new dangers posed by modern algorithmic trading. Multiple asset classes were caught up in the turmoil of a 7% drop in the S & P 500 in just 15 minutes (see first chart below). A number of other asset classes were simultaneously pulled down with equities. Algo trading had actually been increasing for years in advance of the flash crash, and was already taking a 75% share of all trading volumes, as shown in the second chart below.

Multi-Asset Flash Crash 2010:

Source: Reuters; econmatters.com

Source: en.wikipedia.org

Additional flash crashes of note have followed the one in 2010, including a January 6, 2014 gold futures crash (see first chart below), an October 15, 2014 Treasury bond yield crash (see second chart below), a July 18, 2015 bitcoin flash crash (see third chart below), and an August 24, 2015 stock market flash crash (see fourth chart below).

Gold Futures Flash Crash 2014:

Source: forbes.com

Treasury Bond Yield Flash Crash 2014:

Source: zerohedge.com

Bitcoin Flash Crash 2015:

Source: forexnews.com

S & P 500 Flash Crash 2015:

Source: Investopedia.com

There are several ways that algo trading operates to secure advantages to institutional investors at the expense of less sophisticated institutional and retail investors. First of course is High Frequency Trading (HFT), which uses front running technology to see incoming trades before they get executed, as shown in the first chart below. HFT traders can then (legally) front run the trade and pocket a small gain in a few milliseconds. This HFT trading ecosystem was subjected to a massive dose of scrutiny when famous author and analyst Michael Lewis published his expose entitled "Flash Boys" (2014; W.W. Norton & Co., New York, 274p). Another relatively unfair aspect of HFT is that it allows traders to collect fees from brokers for making a market in the traded stock, as shown in the second chart below. Another common practice is "spoofing," in which a trader makes a fake offer and then withdraws it and takes the other side of the trade, as shown in the third chart below. Spoofing can be spotted on detailed charts of trading activity, as shown on the fourth chart below, and is now a major part of market activity.

Source: trade2win.com

Source: nakedcapitalism.com

Source: blogs.wsj.com

Source: wsj.com

The relative importance of HFT may have peaked in 2009 on an annualized basis (see chart below), but this smooths out some very rough data. In detail, HFT rises to 70% of all trades (or more) on some days, and that is what we should be concerned about here. It is encouraging in a way that the same chart shows revenue for HFT firms falling, but if we assume that this is the usual arbitraging away of advantage as competition arises, then it implies that there are either more players playing the game, not fewer, over time, or there are smaller profits to be gleaned.

Annualized HFT Market Share Has Peaked, But Is Still High:

Source: zacharydavid.com

The effects of the HFT ecosystem include increased intra-day volatility (see first chart below), and increasingly bold market manipulation by firms and individuals gaming parts of the system with sophisticated technology (see second chart below).

Increased Volatility Due to HFT:

Source: valuewalk.com

Suspicious Activity Reports (Futures) Increasing With Time:

Source: securitiesanalytics.com

Yet another market innovation has been the creation of so-called "dark pools," private exchanges or forums for trading securities that generally involve institutional block trades (large volumes of shares submitted at one time). The key reason for these private pools is the confidentiality they offer, which helps large institutions avoid showing their hand when trading large volumes of a security, and limits the market impact of the trades because the size and identity of the trading firm are hidden from view. This has been a powerful force for change in the markets, and has probably helped cut (reported) trade volumes over the years (see first chart below). Dark pools have become increasingly important in the markets over the last few years, although they may have recently stabilized with a 60% share of the US market (see second chart below). Of course, those outside the dark pools are disadvantaged because they can't really see what is going on, while dark pool players can often see substantially more.

Source: businessinsider.com

The Rise of Dark Pools:

Source: economist.com

A champion of the beleaguered masses (and institutions wary of front running) has arisen in the form of a competitive new exchange (NYSE:IEX), the heroes of Michael Lewis's book, led by Brad Katsuyama. They have built a successful alternative to HFT-dominated exchanges like NASDAQ (NADQ). They are focused on removing the unfair advantages of the HFT ecosystem, so much so that they have built in a significant (350 microsecond) delay (speed bump) in their trading system to prevent front running. Their story is interesting, but their success so far is also encouraging, as shown for the first few months of their operations in 2014 in the chart below; they are now about to see if they can grab a major share of the markets, according to various Seeking Alpha authors.

IEX Exchange Volumes in First Months of Operations:

Source: financetrainingcourse.com

In conclusion, there is strong evidence that the HFT ecosystem has increased the risk in the system. Dark pools have decreased the transparency in the system. Flash crashes are frequent and regulators have done very little to stem their flow. As we saw in 1987, and as some of the more gloomy analysts have noted, also in 1929 (see John Hussman's chart below), market conditions prior to a crash show evidence of stress in the system, and can provide warning signs that all is not well. A combination of multiple negative forcing factors causes crashes, and just like in most airplane crashes or in the movie "Dr. Strangelove," a series of small errors, taken in combination, can deliver utter catastrophe. But you know, retail investors have a choice: they can assume that tail risk is large and plan accordingly, or they can do what Slim Pickens did at the end of the movie (see photo below): they can enjoy the ride.

Topping Pattern and Subsequent 1929 Market Crash:

Source: John Hussman, hussmanfunds.com

Actor Slim Pickens Takes A Ride on a Nuke in Dr. Strangelove:

Source: tigersweat.com

Prudent investors should keep an eye on the VIX Index, on the SKEW Index, and on market breadth indicators like New Highs Minus New Lows in order to spot deterioration in market internals over the next few weeks or months. It might make sense to at least stay away from exchanges like NASDAQ Inc. (NADQ), Intercontinental Exchange Inc. (NYSE:ICE), or CME Group, Inc. (NASDAQ:CME). Allocations might be best put in defensive mode at this juncture, given high valuations, poor liquidity in bond markets (at least), skittish investor psychology, rampant regional markets arbitrage, and periodic flash crashes. It would make sense for such investors to move to relatively high allocations to bonds (Vanguard Extended Duration Treasury ETF EDV, Wasatch-Hoisington US Treasury Fund WHOSX, I-Shares 20+ Yr. Treasury Bond ETF TLT, I-Shares 10-20 Yr. Treasury Bond ETF TLH, I-Shares 7-10 Yr. Treasury Bond ETF IEF, or SSGA Active ETF SPDR DoubleLine Tactical Bond TOTL) and cash. Equity allocations could include some long/short funds such as Diamond Hill Long/Short Fund (MUTF:DHLSX), Otter Creek Long/Short Opportunity Fund (MUTF:OTCRX), or AQR Long/Short Equity Fund (MUTF:QLENX).

Disclosure: I am/we are long TLT, TLH, IEF, TOTL, OTCRX, QLENX.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: This article is intended to provide information to interested parties. As I have no knowledge of individual investor circumstances, goals, and/or portfolio concentration or diversification, readers are expected to complete their own due diligence before purchasing any stocks or other securities mentioned or recommended.