Rebalancing

To see this article in digital magazine format, click here.  

“People set rules,” Duke University’s men’s basketball coach Mike Kryzewski once said, “to keep from making decisions.” For the most part, this is the mantra of the financial industry, be it high-flying mixed-strategy hedge funds, or anti-regulatory lobbyists who, like Kryzewski, think that “too many rules get in the way of leadership… and leadership is ongoing, adjustable, flexible, and dynamic.”

However, in the world of institutional and mutual fund investing, that lack of flexibility is sometimes exactly the point. Pension funds aren’t supposed to be playing it fast and loose with hard-earned retirement money. In the case of private retirement funds like 4Oi(k)s, neither the government giving out tax breaks nor the people who plan on retiring before they’re 90 want their futures riding entirely on the whims of some Wall Street trader they don’t know. After watching one-time investment darlings like Harvard University reduced to serving its students the equivalent of economy-hotel continental breakfasts after losing 30% of its endowment in 2008/2009, restrictions and caution are becoming the name of the game in even the more flexible endowment world.

One of the most rigid and time-honored rules for many funds is maintained through portfolio rebalancing. Funds create a predetermined asset-allocation mix, with a fixed percentage of total assets invested in different classes. A standard (or, perhaps, theoretical) mix for most funds is 60% invested in equities and 40% in fixed income, a (rough, depending on which markets are up and down) reflection of the U.S. market’s overall mix of stocks and bonds. Then, managers either revisit that mix at fixed intervals (monthly, quarterly, yearly) and buy and sell to bring it back into line, or create triggers—a fixed range, usually within 5% to 10% of the original mix for each asset class— outside of which the portfolio has to be rebalanced.

The goal of a balanced portfolio is not only diversification, but also predictability, and the security that many investors assume comes with it. For many funds, says Arun Muralidhar, former head of research for the World Bank’s pension fund and the Chairman and Co-CIO of AlphaEngine Global Investment Solutions, “Consultants help them craft investment statements for good governance reasons, to appeal to investors.”

However, in the investment world, predictability also means inefficiency. An entire stratum of investment strategies has cropped up over the years to take advantage of the regular rebalancing of big pension and mutual funds (and, in a similar vein, index funds updating their mix and weighting of stocks). Hedge fund legend Michael Steinhardt made millions by taking huge block trades off the hands of large funds at a sizable discount from the market. Since the funds were looking to unload the equities simply for the purpose of rebalancing, not based on new information, traders like Steinhardt could collect a handsome profit by simply selling them off in small enough increments as to not spook the market. In more recent years, pairs-trading quant funds have looked for and exploited tiny market deviations, often caused by big selloffs and purchases. High-frequency traders are similarly content to take steady profits by nibbling around the edges of these massive trades, like pilot fish trailing a shark and picking up tiny morsels from his big kills. There are even service industries aimed at combating these pilot fish, like “dark pool” exchanges and algorithmic trading programs that try to minimize the market impact of large trades.

Yet, when it comes to potential problems with regular portfolio rebalancing, there are bigger fish to fry than the couple of basis points lost to inefficiency: a more fundamental irony at the heart of the very idea of portfolio rebalancing that’s being questioned by academics and big funds alike. As Nobel Prize winner and Emeritus Professor of Finance at Stanford’s Graduate School of Business William Sharpe pointed out in a paper last fall, “Rebalancing a portfolio to a previously-set asset-allocation policy involves selling relative winners and buying relative losers.” Byway of example, let’s say you manage a $100 million fund with the rudimentary 60/40 mix. Let’s say that in the course of a year the stocks perform well, with a 20% return, while bonds bring back just 4%. That turns your $60 million in stocks into $72 million, and the bonds from $40 million to $42 million, transforming your 60-40 mix into a 63-37 ratio. So, to rebalance, you’ll be forced to sell off $3.6 million worth of your high-performing stocks, and buy up the same amount’s worth of poor-performing bonds (minus, of course, the not-insignificant transaction costs). It’s what Sharpe calls a “contrarian strategy”—or what others might call throwing good money after bad.

There are, of course, some reasonable justifications for rote rebalancing. If the portfolio manager thinks that markets often are inefficient, she might see a strong-performing asset class as one that’s perhaps overvalued, possibly even heading toward a bubble. This would mean the poor-performing class was temporarily undervalued and, in the case of something like government bonds, would probably turn into a safe haven should the bubble burst.

This philosophy shares some similarities with George Soros’ Theory of Reflexivity, which holds that investors often are irrational and tend to follow positive investment trends for longer than they should, and panic too quickly and for too long when the trends reverse themselves. Soros, of course, tries to ride the irrational euphoria for as long as possible before moving out of overheated markets but, for a more conservative mutual or pension fund, regular portfolio rebalancing serves as something of an automatic check on the irrational exuberance of trend investing—trimming some big gains, but also cutting risk during the eventual downturns. After all, even Soros can get wrapped up in overvalued assets, as happened to his and Stan Druckenmiller’s Quantum Fund during the 1990s’ Internet bubble.

As Sharpe points out, there is also an efficient market justification for contrarian rebalancing. In an efficient world, the contrarian strategy is irrational for most investors because it sells winners (who, according to the Efficient Markets Hypothesis, must be actual winners, and not illusory bubble-riders, because the markets say they are). However, in the case of pension and mutual funds—and, increasingly, even endowments hoping to avoid Harvard-esque collapses—investors have very specific needs that are different from the investing public at large. They’re hoping for good returns, but demanding stability, and by keeping a fairly consistently diversified portfolio through regular rebalancing, a fund won’t be so prone to the kind of imbalances that can lead to dizzying ups and downs.

However, as AlphaEngine’s Muralidhar points out, the decision to stay slavishly beholden to a predetermined asset allocation often is not based on any kind of explicit investing strategy but is simply a by-product of marketing decisions that can have an unintentional handcuffing effect on portfolio managers. When markets go bad, the seemingly conservative route of regular rebalancing can actually be more disastrous then letting the asset allocation drift. Take for example Yale’s David Swensen, a steadfast proponent of regular rebalancing who lost almost as much money as Harvard during the financial crisis.

This handcuffing is in some ways an intentional feature of many funds, a way to keep traders from making too many potentially risky decisions with people’s retirement funds. But buried at the heart of the option-curtailing is a problematic assumption: that the original decision as to what the asset allocation should be was so wise in the first place, or that changing economic and investing environments don’t call for a similar change in asset allocation.

This last point bears repeating. The general idea behind the conventional 60/40 mix is to reflect the total value of stocks and bonds in the US market, and thus reflect the general risk in the market as a whole. However, as markets fluctuate, that ratio changes. In the spring of 2000, stocks made up as much as 75% of the combined stock and bond markets, so a 60/40 portfolio was weighted toward bonds and, thus, less risky than the market as a whole. Last year, stocks reached a low of 43% of the market, meaning a 60/40 mix would have been considerably riskier than the market as a whole.

In recent years, a number of funds have tried to build more flexibility into their strategies by switching from the old industry standard of reviewing and updating the asset-allocation mix every three years to annually. Tripling your rate of review is nothing to sneeze at, but it’s a bit like going from planning out your daily wardrobe once a year to doing it every few months. You might be able to adjust better to the changing seasons but, on a day-to-day basis, you’re still running the risk of wearing a light sweater in the middle of a snowstorm.

In fact, annually updating asset allocations actually can turn out to be riskier than only doing so every few years. Imagine you did an asset-allocation study at the end of each calendar year. Stocks were riding high in 2007, but would have dropped like a rock in the days and weeks following your end-of-the-year review, finishing 2008 down more than 4500 points on the Dow Jones Industrial Average. That likely would have led to a move away from equities in the end-of-2008 review, a shift made just in time to miss out on the market’s nearly 20% recovery in 2009. It’s a potentially unhappy medium—enough flexibility to lose sight of longer-term goals, but not enough flexibility to truly take advantage of changing market conditions from quarter to quarter and week to week. Meanwhile, some real asset-allocation flexibility and good timing would have served any fund well throughout 2008 and 2009, or during 2010’s spring/summer of discontent and bullish (well, heiferish) autumn. That’s where a bit of short-term tactical asset allocation can supplement longer-term strategic asset allocation. By recommending investment ranges (say, 50% to 70% invested in stocks) instead of a single target, money managers can have the freedom to, well, manage the money based on what’s happening in the markets, instead of conforming to pre-set conditions. Most funds already do this, of course, but academics such as William Sharpe are looking further into the issue, urging large institutional investors to reconsider their asset allocation and portfolio rebalancing strategies. In a recent presentation to CalPERS, the largest public pension fund in the country, MIT’s quantitative finance scholar Andrew Lo urged the fund’s board to revisit its staid risk-management and asset-allocation policies. “When the environment becomes unstable, then it’s the height of irresponsibility to keep a static portfolio,” Lo told the fund, according to Bloomberg. “This notion of tactical risk management is going to become more important than ever before.”

One study, by Martin Leibowitz and Anthony Bova of Morgan Stanley, found that, over a 10-year period, portfolios that were allowed to drift actually performed better than those that were held to their asset-allocation targets. Another study by institutional investing consultants Wurts & Associates found that, since 1926, a more flexible form of rebalancing not only outperformed the standard Strategic Asset Allocation form of rebalancing on average, but also actually lessened the impact of drawdowns. Others, like Princeton’s John Mulvey, have found ways to keep the “buy low, sell high” elements of regular rebalancing, without the slavish vigilance and high transaction costs that it can involve.

So, what is the forward-looking chief investment officer to do? The options are myriad. Sharpe recommends two alternate strategies for portfolio rebalancing. The first is what he calls “optimization based on reverse optimization,” which is about as complicated as it sounds and requires a fair bit of expensive computer modeling. The second is a slimmed-down “adaptive asset allocation” policy that essentially adjusts a fund’s mix of stocks and bonds to reflect changes in the overall ratio of stock and bond values in the U.S. market.

AlphaEngine lets clients devise a program that Muralidhar calls “SMART Rebalancing” (which stands for Systematic Management of Assets using a Rules-based Technique). By modeling any number of factors (P/E ratios, macroeconomic factors, the width of a CEO’s ties), SMART rebalancing actively manages a portfolio’s beta and also can function as a tail risk hedge, helping lessen big drawdowns during bad times, and without the hefty fees of many tail risk hedging (aka, cover your ass-et hedging) products being offered on the market.

“Our clients are becoming much more interested [in more dynamic rebalancing], says Scott Whalen, Executive Vice President of Wurts & Associates. “[Wurts] looks at valuations on a quarterly basis and as things get out of whack, because markets aren’t always efficient, we make recommendations. In 2009, we were recommending funds go out of their usual range to invest more in equities.… more and more funds are getting bigger acceptable asset allocation ranges from their boards so, when a fat pitch like equities in 2009 comes along, they can move outside of their usual range.”

However, while there’s a broad range of ideas, strategies, and products being offered to help institutional investors manage their asset-allocation mix, on nearly all fronts, the message seems to be the same: creating an asset allocation mix and then never thinking about it again, other than to rebalance unthinkingly, is probably a thing of the past. That’s not to say that a static mix and slavish adherence to it, even as the market changes, is always a bad thing. Take the earlier example of a closely followed 60/40 mix, which would have meant a portfolio that was safer and more bond-heavy than the market as a whole in 2000, and riskier and more equity-weighted in early 2009. In both cases, the static approach actually would have worked quite well, putting money into bonds before the tech bubble burst, and into stocks just before the market’s rally over the last year and a half. Yet, just because a broken watch is right twice a day doesn’t mean you should wear one.