Not long ago, fatigued investors were looking for alternatives to equities.
Shorting stocks became more popular for investors who embraced the idea that making money in a sideways market required stock picking unconstrained by a long-only approach. As usual, the investment industry was quick to capitalize on this popular theme by rolling out products to meet the demand for market-neutral investments.
Traditional equity managers also started moving in this direction. In April 2012, an industry-wide amendment permitted mutual funds to short up to 20% of a fund’s net asset value. Removing the constraint on shorting amplifies stock picking and, in a sense, puts traditional active management on steroids. This is worthwhile if the underlying process adds value.
But a large body of literature claims traditional active management — or, in this case, stock picking — fails to add value after fees and expenses. This challenges the view that magnifying the impact of stock picking by removing the short-selling constraint offers a tangible benefit to investors.
Studies of traditional active management have mostly been limited to long-only mutual funds because the data set is larger and far more reliable than that for alternative investments. (Hedge fund databases can lack integrity because manager reporting is voluntary, which often results in survivorship bias, self-selection bias and backfill bias in the data.)
Despite the long-only nature of the mutual fund literature, investors should not dismiss the relevance of the findings to managers who also go short.
One of the main reasons for removing the short-selling constraint is to enable managers to capitalize on their ability to identify over-valued securities. But if this skill were reliable and systematic, you would expect to see it in the multitude of tests on the long-only universe of mutual funds.
Although managers in this universe cannot short the stocks they believe are set to decline in value, they can choose not to own them, which would result in persistent outperformance relative to their long-only benchmarks.
In simple terms, being able to short stocks may double the stock-picking opportunity set, but that offers no benefit if traditional stock picking doesn’t work to begin with.
Since there is no value from simply increasing the range of stock-picking opportunities by allowing managers to short, a market-neutral strategy will have an expected return of T-bills minus fees and expenses, where the standard is “two and 20.” This strategy holds unless the portfolio is long in another dimension of expected return that is persistent and pervasive — such as size, relative price or direct profitability.
Say a market-neutral strategy with systematic exposure to one or more of these dimensions is included in a portfolio that already has market beta. The client would be better off reducing the “two and 20” and eliminating the cost of shorting by combining market beta with the desired exposure to other dimensions of expected return in one long-only strategy.
A market-neutral approach
We’ve already addressed the error of assuming that good stock picking is the answer. But there’s a bigger problem with adopting a market-neutral approach as a solution for the sideways market, or one where the expected return is zero. The problem is the notion is a fallacy, because there’s always a positive expected return on capital.
That doesn’t mean your return is guaranteed to be positive, but it’s always expected to be. No return is guaranteed because the market can only incorporate what is knowable — and unknowable information is, by definition, new information. If the information were considered bad news, or if risk or risk aversion were to increase and investors were to require higher expected returns, prices would drop.
The market mechanism brings prices to equilibrium where, based on the new information, the expected return on capital remains positive and commensurate with the level of risk or risk aversion in the market.
The opposite would be true if the new information were good news, or if risk or risk aversion were to decline. This is how well-functioning capital markets result in persistently positive expected returns. Investors shouldn’t expect a sideways market ex-ante.
Good and bad news also have to be defined relative to expectations, rather than in absolute terms. The new information could be considered bad news in an absolute sense: for instance unemployment increasing to 9%, but this information could be received as good news in a relative sense if market prices reflected expectations of an increase to 10% unemployment (see “How new info affects stock prices,”).
Investors betting on a sideways market are obviously expecting a below-normal return on equities. They aren’t betting on the future being good or bad, but worse than expected.
Nobel laureate Friedrich Hayek described the market as a communication mechanism that “garners, comprehends and disseminates widely dispersed information better and faster than any system man has deliberately designed.”
When you consider that expectations are already embedded in prices, betting against them is short-sighted.
How new info affects stock prices
|New Information||Relative to Expectations||Result|
|Good news||Better than expected||↑||Above normal return|
|Good news||As expected||—||Normal return|
|Good news||Worse than expected||↓||Below normal return|
|Bad news||Better than expected||↑||Above normal return|
|Bad news||As expected||—||Normal return|
|Bad news||Worse than expected||↓||Below normal return|
Brad Steiman is director and head of Canadian Financial Advisor Services and vice president of Dimensional Fund Advisors Canada ULC.
Originally published in Advisor's Edge Report