A standalone short book is structurally disadvantaged. When paired with a long portfolio and actively rebalanced, shorting transforms into a source of incremental return and risk reduction.
Shorting has an intuitive appeal. If some stocks fall, why not profit from that? The problem, as this paper shows, is that shorting is not just the opposite of buying stocks, and the portfolio structure matters a great deal.
In his May 26, 2026 article “Shorting and Rebalancing: How shorting can complement a long book,” (which picked up a thread from his earlier note on short selling), Verdad’s Greg Obershain asked the simple but important question: does shorting work well on its own, or is it more useful when paired with a long book and actively rebalanced?
To answer the question Obenshain examines two distinct questions. First, why is a standalone short-dedicated fund structurally difficult to operate, even when the manager has genuine skill at identifying losing stocks? Second, how does that picture change when the short book is paired with a long portfolio and the two are rebalanced together?
His analysis draws on historical data covering the 2,000 largest US stocks from 1990 to 2026, constructing a hypothetical fund — nicknamed “Lookahead Capital” — with unusually good short-selection ability, identifying stocks from the bottom 60% of the return distribution. He then compared that standalone short portfolio with a long–short portfolio that combined a long book and a negatively correlated short book. By comparing the standalone short results to long-short combined results, the paper isolates the precise mechanism by which shorting adds value (or fails to).
The article begins with an often-overlooked but important fact: stock returns are asymmetric—stocks can rise without a practical ceiling, but they can fall only 100%, so the distribution of outcomes is not symmetric.
Key findings
· Most stocks go up, but nearly 40% lose money in any given year. Since 1990, roughly 60% of the largest 2,000 US stocks have had positive returns over a 12-month horizon, leaving a substantial pool of losing stocks. On the surface this seems like a rich opportunity for short sellers.
· Shorting has an asymmetry problem that destroys return. When returns are flipped to simulate a short position, realized gains are significantly lower than the raw average of the negative returns. Lookahead Capital’s short book, which had an average long return of –9.0%, translated into only +5.8% when shorted — a 3.2 percentage-point leakage. The culprit is the mathematical asymmetry of gains and losses: a stock that rises 25% and then falls 20% nets zero for the long holder, but nets –10% for the short holder over the same sequence.
· Rebalancing forces a short book to “buy high and sell low.” Maintaining a short portfolio at constant market weights requires covering shorts when a stock rises (buying high) and adding new shorts when a stock falls (selling low). This mechanical rebalancing is the direct source of the return drag. Inverse ETFs that must track an index experience exactly this problem.
· A long-short structure with rebalancing reverses this dynamic. When a short book is paired with a long portfolio and both are rebalanced to target weights, the negative correlation between them becomes an asset. When the long book wins and the short book loses, rebalancing trims the long (selling high) and adds to the short (buying low). When the short book wins and the long loses, rebalancing does the opposite. The combined portfolio in the analysis delivered an estimated realized return of 18.7%, compared to simply adding the two books’ standalone returns of 16.6% — an extra 2.1 percentage points from the rebalancing premium.
· The long-short structure is also capital efficient and reduces volatility. Because longs serve as collateral, the short book requires no additional equity. And the negative correlation between books lowers the overall portfolio standard deviation, reducing drawdowns without necessarily sacrificing return.
The mechanics behind the Finding
· The core insight is that volatility drag works against a short-only investor, but for a long-short investor. The standard estimate of volatility drag is approximately -0.5 x (standard deviation)2. For a standalone short book, this drag compounds on top of the inherent upward drift of equity markets. For a long-short portfolio with negative correlation between the two books, rebalancing harvest a return premium that more than offsets the drag.
· Fund families (such as AQR) that specialize in shorting have learned to work around the structural problem by operating through managed accounts that allow proper rebalancing, or by explicitly running accounts that combine long and short positions together. The key in both cases is that the negative correlation of returns is put to active use through periodic rebalancing.
Implementation: Like vs. Unlike Securities
A related implementation point deserves emphasis. It is important for investors to distinguish between going long and short like securities — for example, long Coca-Cola and short Pepsi-Cola — versus going long and short unrelated securities, such as long Citibank and short Newmont Mining. In the first case, the two positions are drawn from the same industry and tend to share common risk factors, so adding the short reduces net exposure and lowers risk. In the second case, the two securities operate in entirely different industries (banking versus gold mining), move on different drivers, and carry no natural offset. The investor is effectively making two separate directional bets with no offsetting correlation, which is risk-enhancing rather than risk-reducing.
Pairing like securities also reframes the investment thesis. The bet becomes a relative value trade — Coca-Cola simply needs to outperform Pepsi-Cola, regardless of what either stock does in absolute terms. That is a fundamentally different mindset than hoping the long book rises and the short book falls, and it is generally a higher-conviction, lower-variance way to express a view.
This distinction also has implications for how investors should interpret gross leverage. When long and short positions are paired in like securities, gross leverage overstates economic risk, because the two legs offset much of each other’s exposure. Gross leverage is therefore a poor proxy for actual risk in a like-for-like portfolio, and investors should look instead at net exposure and factor-level risk decomposition.
Investor takeaways