The challenge with this risk-management style was that it cuts against the grain of almost every known cognitive bias. Risk aversion tells a trader to sell a winning trade early for fear of losing her gains. Anchoring causes traders to stick with a losing trade based on belief in the original thesis, even after losses pile up. Confirmation bias causes traders to ignore incoming information that calls a thesis into question. A range of cognitive biases lumped under the heading of denial, including the ostrich effect, postpurchase rationalization, and selective perception, cause traders to ignore losses with the comment, “Don’t worry, the market will come back.”
Other biases cause traders to buy securities after price run-ups have already occurred. These come under the headings of availability heuristic and attention bias; the trader is attracted to a security because it receives ample attention in the media due to recent outperformance. This often leads to a poor entry point for the trader. Kovner was categorical on the topic of entry and exit points. He said a proper investment thesis was only half of what was required to make money. The other half was getting the entry point right. The combination of chasing momentum on the entry point and rationalizing losses on the downdrafts leads most investors to a buy-high, sell-low dynamic guaranteed to lose money.
Kovner had a simple solution for these and other behavioral quirks. You either followed his rules or you were fired. Kovner kept written trading authorizations in his own name on file with all of Caxton’s brokers. A trader might have authority to buy and sell securities with the broker, but as soon as the rules were broken Kovner could override his traders and close out the trade directly with the broker. By the time the trader found out, he might already be on his way out the door.
Some traders lost money, followed the rules, and obeyed the stop-loss limits, yet remained convinced their trading thesis was a winner. These traders could appeal to Kovner for another chance to reenter the trade. Kovner insisted that the trader take a time-out, compose a written thesis on why the trade was a potential winner, and then meet personally with him to discuss the potential for reentry. In practice, a good night’s sleep after exiting a losing trade was usually enough to convince the trader he should move on to a new idea.
Kovner and his ilk are rare exceptions. Most traders cannot overcome cognitive biases and do fall victim to the buy-high, sell-low trap of the momentum trader or the early-sale trap of the risk averse. Kovner beat the market for decades with skill, not luck. He is living proof it’s possible. Still, most traders are not Bruce Kovner. On the whole, active investors do not outperform passive investors, hence the allure of indexing. Cognitive bias is part of the reason, yet an even bigger part is skew.
In this context, skew refers to the fact that a large percentage of total returns in broad-based stock indices are attributable to a small percentage of the stocks in the index. For passive managers who buy the index, skewness does not matter. They own the handful of big winners along with numerous small winners and losers and will exactly match the index return; that’s the whole idea of indexing.
For active managers, skewness can be fatal. If you happen to pick the winners, good for you; you’ll show outperformance. Still, active managers constructing portfolios based on subsets of the index components will miss big winners more often than not, because the winners are so few. The handful of big winners that drive index returns are like a needle in a haystack. If fifty active managers each grab a handful from the haystack, only one has the needle; the rest just have hay.
It’s not impossible for all of the active managers to find the needle in the haystack; it’s just extremely unlikely because the needle—the winning stock—often soars based on some unexpected news or exogenous shock that no amount of fundamental research will reveal. This harks back to Merton’s emphasis on inside information.
A recent research paper succinctly describes the impact of skew:
To illustrate the idea, consider an index of five securities, four of which … will return 10% over the relevant period and one of which will return 50%.3 Suppose that active managers choose portfolios of one or two securities and that they equally weight each investment. There are 15 possible one or two security “portfolios.” Of these 15, 10 will earn returns of 10%, because they will include only the 10% securities. Just 5 of the 15 portfolios will include the 50% winner, earning 30% if part of a two-security portfolio and 50% if it is the single security in a one-security portfolio. The mean average return for all possible actively managed portfolios will be 18%, while the median portfolio of all possible one- and two-stock portfolios will earn 10%. The equally weighted index of all 5 securities will earn 18%. Thus, in this example, the average active-management return will be the same as the index … but two-thirds of the actively managed portfolios will underperform the index because they will omit the 50% winner (emphasis added).
There is no better illustration of the skewness effect than the FAANG stock mania of 2016 to 2018. FAANG is an acronym for Facebook, Apple, Amazon, Netflix, and Google. From January 1, 2016, to March 1, 2018, the FAANG stocks outperformed the S&P 500 by over 50 percent. The FAANG stocks accounted for more than 30 percent of the S&P 500’s total gains over the same period. An active manager who owned 495 of the 500 stocks in the S&P 500, or a representative subset, but did not own the five FAANG stocks, would have underperformed a passive index manager using the S&P 500 benchmark by 30 percent.
The evidence is that active managers can beat the market, yet few do so because of behavioral bias and skewness. Those few active managers who can beat the market, such as Bruce Kovner, tend to have high fees or closed funds, or retire early to manage private portfolios as family offices. This leaves everyday investors with few options other than to join the crowd in index funds. Still and all, the zombielike march of investors into passive indexing has created an entirely new set of dangers that are little understood and will prove to be far more destructive to wealth than active-manager underperformance. In the move to indexing, one kind of risk has been traded for another, and the new risk—hypersynchronicity—is the most dangerous of all.
The Everything Bubble
The demise of active investing and rise of indexing have created a positive feedback loop that insures the next stock market crash will be the greatest ever. Professionals understand this, but don’t care; they make good money in the meantime on commissions or wealth management wrap fees, and invest their own money in ways quite different from those they advise clients. I have yet to meet a hedge-fund billionaire, and I’ve met many, who does not have a large personal allocation to physical gold. They are ready for what’s coming. Their clients are not.
The positive feedback loop dynamic is straightforward. Index managers are desperate to match their benchmark index and may lose their jobs, or at least assets under management, if they fail. At some point in a bull market, a small set of stocks or a particular sector may begin to outperform the index as a whole. The FAANG stocks are a good example, but not the only one. The Nifty Fifty stocks of the late 1960s, and the dot-com stocks of the late 1990s are other examples. The reasons for the initial outperformance are irrelevant for purposes of studying the dynamic. Reasons could be fundamental, based on growth prospects. They could be technical, based on chart patterns. More likely there is no discernable reason. That’s typical of an emergent property in a complex dynamic system; events just happen.
Once the dynamic outperformance begins, investors have to buy more of those stocks in order to rebalance portfolios and keep up with the index. The added buying tends to drive up the price and leads to further outperformance. This leads to further buying and further price increases. The dynamic continues like a cat chasing its tail, with more buying, higher prices, more buying to keep up with the index, and higher prices still.
Eventually the positive feedback loop turns negative. This happened with the Nifty Fifty in the 1973–1974 stock market crash, a 45 percent decline measured on the Dow Jones Industrial Average, and with the dot-com crash
in 2000–2002, a 78 percent decline measured on the NASDAQ Composite, both popular index benchmarks. It happened with the FAANG stocks in late 2018. A bull market feedback loop can operate for years before its phase transition to a bear market feedback loop. The changed direction in the feedback loop emerges suddenly, unexpectedly, and often for no evident reason, although there is never a shortage of just-so stories to ostensibly explain what occurred.
The emergence of an outperforming stock sector in recent years and the rise of asset valuations generally are products of former Fed chair Ben Bernanke’s seven-year experiment with zero interest rates and quantitative easing, continued for a time by his successor, Janet Yellen, from 2008 to 2015. Bernanke relied on what he called the portfolio channel effect. The idea is that if the Fed holds short-term rates at zero, and depresses long-term interest rates by buying Treasury notes, investors are forced to look elsewhere for higher yields. In doing so, investors will bid up the prices of stocks, corporate bonds, real estate, emerging markets, and other assets. The resulting gains in asset prices will provide collateral for loans to corporations and boost consumer confidence as the gains show up in 401(k) statements. This newfound wealth and confidence would gin up spending and more lending. The combination of corporate borrowing, investment, and consumer spending would soon set the U.S. economy back on the path of self-sustaining trend growth.
Bernanke’s experiment failed. Investment and consumption did not return to trend. Average growth in the U.S. economy in the nine years after the end of the recession in June 2009 was 2.2 percent, far below long-term trend growth, and the weakest recovery in U.S. history. The Fed’s balance sheet was leveraged over 120:1 and stuffed with $4.5 trillion of bonds that left it ill-prepared to deal with a new recession, should one arrive.
Like various failed experiments, there were noxious by-products. Debt issuance by governments and corporations exploded on the back of low rates. Consumers took on $1.6 trillion of student loans. Skyrocketing default rates on student loans damaged credit ratings of graduates and parent cosigners, which impeded hiring and household formation and healthy consumption patterns that go with each. The most poisonous side effect was the inflation of asset values into what observers call the “everything bubble.”
Evidence for an everything bubble is abundant. Robert Shiller, winner of the Nobel Prize in economics in 2013, made this observation in a 2015 interview:
I define a bubble as a social epidemic that involves extravagant expectations for the future.4 Today, there is certainly a social and psychological phenomenon of people observing past price increases and thinking that they might keep going. So there is a bubble element to what we see. But I’m not sure that the current situation is a classic bubble …. In fact, the current environment may be driven more by fear than by a sense of a new era …. This time around, bonds and, increasingly, real estate also look overvalued. This is different from other overvaluation periods such as 1929, when the stock market was very overvalued, but the bond and housing markets for the most part weren’t. It’s an interesting phenomenon.
Shiller’s surmise is that bubble dynamics are emerging not because of pie-in-the-sky hopes of stock investors, but for fear of missing out on gains that might be available on any risky asset class in a world of zero interest on safe assets. This fear is driven by a desperate attempt to rebuild lost savings destroyed in the 2008 global financial crisis. Bubble dynamics appear in stocks, bonds, real estate, auto loans, student loans, emerging markets, cryptocurrencies, and beyond. This result would come as no surprise to an Austrian School economist. The everything bubble is classic malinvestment—the misallocation of savings—that accompanies easy money.
Between 2010 and 2016, the S&P Case-Shiller Home Price Index for San Francisco rose 68 percent, from 139 to 234. The comparable index for Sydney showed a 69 percent increase, from 98.9 to 167.6. The Canada Real Residential Housing Price Index was comparably bubbly, rising from 100.0 in January 2010 to 143.1 in September 2017, a gain of over 43 percent. Similar gains were seen in housing markets all over the world, from Melbourne to Miami and from London to Los Angeles. Some of this was indigenous demand from Silicon Valley billionaires, but diverse gains were driven by flight capital as Russian oligarchs, Chinese princelings, and Venezuelan elites fled unstable and capricious jurisdictions for safer climes.
The stock market had its frothy fables as well. As of August 2017, Tesla had a market capitalization of over $750,000 per vehicle sold, compared to $16,000 for Toyota and $5,500 for General Motors. The auto sector as a whole was kept afloat on a sea of loans. Between 2010 and 2017, U.S. auto loans outstanding surged from $650 billion to $1.1 trillion, of which $280 billion were rated subprime. In the same period, delinquent auto loans increased by $23 billion. Corporate credit was in no better condition than consumer credit. As of August 2017, U.S. corporate debt outstanding stood at $5.9 trillion, a 54 percent increase from 2010. American dollar-denominated debt issued by emerging-markets companies exceeded $9 trillion by 2017, according to the Bank for International Settlements, or BIS.
These equity and credit bubbles were visible from bank and corporate balance sheets. Behind those was a wall of invisible liabilities in the form of derivatives. The five largest U.S. banks held $157 trillion of derivatives measured by gross notional value at the end of 2017, a 12 percent increase from the comparable amount of derivatives immediately before the 2008 global financial crisis. That increase may seem modest, yet it is contrary to repeated claims by regulators that the financial system is safer and less leveraged; it’s not. Even the 12 percent increase in derivatives exposure since the last crisis is not a complete picture. That figure includes only what banks hold in off-balance-sheet positions. Trillions of dollars of derivatives have been moved out of banks to third-party clearinghouses. These clearinghouses are intended to be another safety valve, because they allow simultaneous netting of multiple derivatives exposures from multiple banks instead of the simple bilateral netting that occurs when banks resort to self-help against failing counterparts in a crisis. That’s helpful when one bank is failing and remaining solvent banks want to liquidate positions quickly. Yet when multiple banks are in danger of failure, as was the case in 1998 and 2008, a clearinghouse is more like a game of musical chairs with no chairs. Like dominoes in a row, as each bank fails the liquidity burden falls on fewer strong hands, until those banks fail also. In that case, the clearinghouse itself is in jeopardy and no longer able to fulfill its functions. Member banks do not record this contingent liability for clearinghouse risk on their balance sheets. Clearinghouses do not eliminate risk, they merely move risk around in ways that make it more difficult to discern. Where derivatives are concerned, the financial system is not smaller, not safer, and not more sound.
While these bubbles grew, a surge in passive investing acted like a force multiplier to malinvestment. Markets have reached the point where indexing itself is a bubble that feeds these individual asset bubbles.
Wall Street never saw a bubble it didn’t like if there was money to be made by inflating it. The index bubble is no exception. Index fund sponsors and passive managers began to crank out easy-to-trade bespoke products that required no active stock selection by investors. The most popular of these were exchange-traded funds, or ETFs, and their close cousins, exchange-traded notes, ETNs. The ETFs and ETNs are technically securities, are registered with the SEC, and trade as listed products on the New York Stock Exchange, NASDAQ, and other exchanges. Each ETF and ETN trades like a single share of stock, producing a simple buy or sell decision for an investor, yet is constructed to have an underlying basket of stocks or notes. For example, an emerging-markets ETF might include equities from companies in Turkey, Brazil, Indonesia, Malaysia, and other developing economies. A retail ETF might include equities from companies in brick-and-mortar store sales such as Walmart, Home Depot, Starbucks, and other similar firms. The possibilities don’t stop with the variety of investment sectors. ETFs can be leveraged so that an invest
or receives three times the return (or three times the losses) of the underlying basket. For example, a two-times health-care ETF returns twice the gains of an underlying basket of health-care-related stocks such as UnitedHealth Group, Medtronic, and Aetna. The assets under management in the leveraged equity ETF sector have grown from $5 billion in late 2007, when the product was invented, to over $30 billion by early 2018. Finally, ETF returns can be inverse to the underlying stocks, so if the selected group goes up, the ETF goes down. Such ETFs leave the purchaser in the position of a put option seller without standard safeguards. To sell put options in a brokerage account requires special account forms, added due diligence by the broker, risk disclosures, and stringent margin requirements. None of that is required with inverse ETFs; you just buy them and take your chances. Leaving no stone unturned, Wall Street naturally offers inverse-leveraged ETFs.
The real danger in ETFs, especially those structured with leveraged and inverse performance, is relative illiquidity. When an investor sells a share of an individual stock such as IBM, they are selling into a relatively deep pool of potential IBM buyers. When investors sell a technology-oriented ETF that may happen to include IBM along with other less-liquid technology stocks, the pool of buyers for that ETF may be quite small, especially in a steep market drawdown, let alone a panic. The ETF could fall faster than some of its components, placing selling pressure on all of the components, as authorized dealers buy the ETF and short the underlying stocks as an arbitrage. Any ban on short selling, as happened in 2008, eliminates the arbitrage and leaves the ETF in free fall. These positive feedback loops (“positive” in the sense of self-reinforcing, not in the sense of desirable) are possible to model theoretically as hypersynchronous events, yet cannot easily be predicted as real-world events. The feedback loops are unforeseen emergent properties of complex dynamic systems; what some call black swans, and they are everywhere.
Aftermath Page 15