Campbell R. Harvey’s presentation of his joint work with Michele Mazzolini and Alessandro Melone examined how large institutional investors’ routine rebalancing rules can quietly move markets and impose hidden costs on pension beneficiaries. He focused on investors such as defined benefit pension funds, target date funds and university endowments that maintain relatively fixed allocations such as 60-40 between equities and fixed income. The central idea is that because these portfolios must periodically be brought back to target weights, their trades become mechanical and therefore predictable. As Harvey put it, his paper is about “the hidden costs of rebalancing,” asking whether, given the size of this “rebalancing complex,” these trades affect overall market prices and create predictable variation in returns that others can exploit.
To study this, he distinguished two main types of rebalancing rules that many funds use and often publicly disclose. One is threshold rebalancing, where a fund only trades when its allocation drifts beyond a fixed band around the target, for example, a 2% band would imply that a 60-40 fund rebalance if equities move above 62%. The other is calendar rebalancing, where funds adjust back to target on a fixed schedule, such as the last day or last week of each month or quarter, regardless of how far the weights have drifted. Both systems were designed for sensible institutional reasons – thresholds reduce transaction costs, calendars minimize tracking error versus benchmarks – but Harvey emphasized that both “induce predictability” in trading flows because outsiders can anticipate when and in which direction the big funds will trade.
Empirically, Harvey built rebalancing “signals” that capture how far portfolios are from their targets and then ran daily predictive regressions of S&P 500 returns minus Treasury bond returns on these signals. He found that when equities were overweight, funds must sell stocks and buy bonds, and this was associated with a negative predicted equity–bond return the next day; when equities are underweight, the opposite happens. He estimated that when stocks are overweight, the selling pressure leads to around a 17 basis point impact on equity returns and about a 3 basis point impact on bond returns, effects that are temporary and dissipate after a couple of weeks. He stressed how robust the findings were: even after adding control variables like momentum, volatility indices (VIX and MOVE), economic uncertainty and sentiment, “the coefficients are barely changed,” and the same basic pattern appears whether he looks at futures or spot markets and when he splits results into equity and fixed-income components.
The presentation also highlighted the front-running dimension. Through a roundtable with CIOs of major public pensions, Harvey learned that many funds were fully aware they were being front-run, and some even allowed their own alpha desks to trade ahead of their and their peer’s rebalancing. He described a simple trading strategy based on his signals that delivers sizable excess returns with positive skew, that is especially profitable in periods of high illiquidity when the market impact of the rebalancing is the greatest. In his view, “the mechanical rebalancing rules do not serve the pensions well,” because they amount to “pre-announcing” large trades to the market.
While he illustrates that randomizing the timing of rebalancing could greatly reduce costs, he is clear that the main contribution of the paper is to document the problem – showing that mechanical rebalancing generates predictable, temporary price pressures that are economically important for pensioners and should motivate a reconsideration of current rebalancing policies.