Portfolio management

The unexpected costs of rebalancing and how to address them

082714 sandy rattray
Man AHL CEO Sandy Rattray

Investors commonly rebalance their portfolios to maintain desired asset allocations as prices change.

Pension funds and other long-term investors generally use such fixed-weight strategies. These strategies have intuitive motivations, but we believe key properties of them are poorly understood by many investors. In particular, rebalancing can magnify drawdowns when there are pronounced divergences in asset performance. Such divergences are usually driven by equities, and in late 2008 and early 2009, some rebalanced strategies underperformed passive strategies by hundreds of basis points.

In traditional 60/40 portfolios, the vast majority of the risk (about 85%) comes from the 60% allocation to equities. Equity indexes, in addition to generally having much higher volatilities than bond indexes, have fat downside tails, i.e. large negative returns, and sharp drawdowns occur more frequently than large positive returns. Rebalancing strategies not only inherit these properties, but actually exacerbate the drawdowns.

A momentum overlay can reduce the risks associated with rebalancing by improving the timing of the rebalance. The rebalanced portfolio in combination with the momentum overlay potentially reduces drawdowns and improves risk properties during crisis periods where stock and bond market returns diverge, while maintaining relatively stable portfolio weights over the long term.

Fixed weight vs. drift weight

A passive buy-and-hold strategy is an alternative to a strategy of rebalancing to fixed weight. The capital allocation within the portfolio will drift as market prices change, but the investor remains passive. In contrast, the active rebalancing strategy will restore the target fixed weights.

The motivations behind a rebalancing strategy are straightforward. First, the drifting weights, implied by the buy-and-hold approach, might become extreme as assets diverge and few investors would remain passive indefinitely. The strategy of rebalancing to fixed weights puts this adjustment process on a regular schedule.

The so-called “rebalancing premium” is a second motivation for the fixed-weights strategy. The premium refers to the extra returns that the rebalancing process generates under certain circumstances. Consider how such profits might arise. Divergent asset performance will cause the weights to differ from the target allocation. To restore the desired allocation, the investor must buy some of the underperforming assets and sell some of the outperformers. Buying low and selling high has an intuitive appeal, and provided there is mean reversion in relative asset performance, rebalancing to fixed weights might generate positive returns.

On the other hand, when assets diverge strongly over time, fixed-weight rebalancing might cause losses relative to buy-and-hold as the rebalancing continues to increase the allocation to the underperforming asset.

The performance difference between fixed weights and drift weights shows the characteristic return profile (sometimes called negative convexity) of an investor who has sold both put and call options (although no actual options are sold of course). This profile shows positive, but limited, returns under small moves and large negative returns under large moves. The driver is the divergence in stock and bond returns because differences in return change the allocation.

Buying losers and selling winners is the essence of rebalancing to fixed weights. Market dynamics ultimately determine whether such a trading strategy is profitable. A trendless environment, especially one characterized by mean reversion, would clearly be friendly to this strategy. However, trending markets, e.g. where equities keep losing relative to bonds, pose a problem for rebalancing. A sustained trend means one buys an asset on the way down. Trending is a well-established property of markets over the last century and remains an important trading signal. One ignores trends at one's own peril.

Rebalancing is therefore no free lunch, but represents an implicit bet against diverging asset performance. The cost occurs when markets trend and rebalanced portfolios experience deeper drawdowns than buy-and-hold portfolios. For small values of stock-bond return divergence, fixed-weight rebalancing modestly outperforms buy-and-hold; for large divergences, the underperformance of the rebalancing strategy is marked. This extra risk is therefore a fundamental property of the dynamic behavior of the rebalanced strategy.

Opposite characteristics

Momentum has the opposite characteristics to fixed-weight rebalancing.

Momentum trading involves the purchase of winners and sale of losers, while maintaining a given volatility exposure. Technically, we can say momentum trading produces a positively convex and positively skewed distribution of returns.

Remarkably, the positive convexity and skewness do not depend on strong assumptions about the underlying markets; but can be proved mathematically. The dynamic style of trading introduces the positive skewness, which translates into reduced drawdowns relative to passive strategies.

The concept of an overlay is familiar to many pension plans, where overlays are commonly used, particularly on the liability side. But on the asset side, momentum overlays can be used to hedge the drawdown risks mechanically induced by rebalancing. The overlay tends to gain from the very trends that hurt fixed-weight rebalancing, but imposes little cost in other environments.

Why don't rebalancing and momentum just cancel?

Doesn't adding momentum to the rebalancing process merely get you back to where you started: a passive buy-and-hold portfolio?

The answer is no. While rebalancing and momentum signals treat outperforming (and underperforming) assets in broadly opposite ways, with rebalancing selling winners and momentum buying them, there are important differences in the trading.

The momentum overlay, operating on a daily time-scale, effectively acts as a timing strategy around the monthly or quarterly rebalance. For example, the overlay will resist buying too aggressively into equities in times of market stress (easing the psychological pressure on the rebalancer). But, as divergences in asset performance stabilize, the momentum overlay will cut out and allow the full rebalance to occur.

The driver behind the worst drawdowns is the fact that rebalancing represents a kind of anti-momentum trading in which one buys underperforming assets and sells outperforming assets. Rebalancing, however, is here to stay. By adding a momentum overlay, investors can help mitigate the risk introduced by the rebalancing process and improve the risk and drawdown properties of the rebalanced portfolio.

Nick Granger is portfolio manager of dimension, Sandy Rattray is CEO and David Zou is quantitative strategist, Man AHL, London. Campbell Harvey is investment strategy adviser, Man Group, London and professor of finance, Duke University in Durham, N.C.