Rebalancing in volatile markets
Extreme market events often trigger rebalancing requirements, particularly in funds that apply strategic asset allocation (SAA) policies. However, implementing a rebalancing policy in whipsawing markets is anything but easy and the potential to lock in losses is substantial, a recent webcast hosted by Fund Business has heard.
Speaking on “Rebalancing Revisited”, Episode 5 in the Investment Implementation Webcast Series, Alistair Barker, Head of Portfolio Construction at AustralianSuper; Greg Cooper, Non-Executive Director at NSW Treasury Corporation (NSW TCorp), Perpetual and Colonial First State; Nader Naeimi, Head of Dynamic Markets at AMP Capital; and Laura Ryan, Head of Research at Ardea debated the implementation challenges of a rebalancing policy framework in extreme conditions.
While on paper the concept is reasonably sound – not least because it forces fund managers to stick to their risk/ return objectives – the idea that rebalancing forces fund managers to sell when markets are expensive and buy when they fall doesn’t always hold. In fact, in highly volatile markets, it is quite easy for rebalancing to achieve the opposite effect.
Rebalancing in volatile markets
“When you get extremely volatile markets, depending on your rebalancing policy time frame, it could well be that you are selling when markets are falling and buying when they are rising, rather than the inverse, which is what rebalancing is supposed to do,” Cooper said. He said that process-oriented rebalancing strategies in particular are at risk of mis-timing the market when volatility is at all time highs.
“Even in listed markets or quasi-listed markets like credit, there is the potential to lose a significant amount of value [when] spreads start to widen and you might have to suddenly start crossing a fairly wide spread in order to sell down one asset class and buy another. That policy might work perfectly well in normal market conditions, but in volatile market conditions it could suddenly end up costing quite a significant amount of money,” Cooper said.
If permanent loss of capital is one risk, the liquidity implications can also be very substantial, particularly when funds find themselves forced to sell their most liquid assets in order to fund the purchase of less liquid assets as part of the rebalancing exercise. At a time when liquidity is invaluable, that is not an ideal policy outcome.
AustralianSuper’s Barker said that stress testing liquidity needed to be a key consideration of any rebalancing framework. “Liquidity in a portfolio can be eroded quite quickly when a market is whipsawing around, and you are undertaking quite a reasonable amount of trading. [Therefore], when designing a rebalancing policy, it is really important to have tested the extent to which it might actually exhaust liquidity and to what extent it might need to be suspended if conditions get beyond a certain amount of stress,” he said.
Managing unlisted exposures
If rebalancing a fully liquid portfolio has its challenges at times, adding unlisted assets into the mix adds a further level of complication. That’s partly because valuing unlisted assets is considerably more difficult than valuing listed assets, but also because of their illiquid nature which by definition makes them less easy to buy and sell.
“The greater the proportion of illiquid assets you have got, the more careful you would have to be actually having a rebalancing strategy in the first place,” Cooper said, adding that funds that applied rebalancing policies should ensure that the rebalancing timeframe matched the time needed to sell assets and generate liquidity. “If there is a mismatch between the two - or certainly if your rebalancing period is shorter than [the period] in which you can draw liquidity from your portfolio, then you are going to have issues,” he said.
In this instance, it is extremely important to remain on the front foot and avoid a situation in which you have no other option but to liquidate these assets. “It’s not the illiquid part of the portfolio that has the pain initially, it tends to be the liquid part that causes the illiquid part to drift up. If you wait until you are forced to rebalance, you have to sell at fire-sale prices, causing a permanent loss of capital,” AMP Capital’s Naeimi said. He added that, until the industry takes to introduce more frequent and transparent methods of valuing illiquid assets, there would need to be additional layers of stress testing analysis on illiquid assets.
At AustralianSuper, which runs a comparatively active approach to asset allocation alongside an unlisted assets portfolio, Barker said the longer term nature of the unlisted asset program is carefully balanced against the flexibility required to maintain the dynamic program, in recognition of the fact that the two effectively compete for liquidity. “There is a trade-off between the flexibility required to run a dynamic process and the need to make longer commitments in unlisted assets. In effect those two things effectively compete for liquidity. [..] You can’t do all of both, [so] you really have to decide where you are on both and then size the programs accordingly,” Barker said.
Alternatives to SAAs
While any strategy with a target allocation would require rebalancing, these challenges are most pronounced in SAAs. Ryan questioned whether - given the difficulty in forecasting returns, the instability in covariance matrices and the long horizon required to achieve return objectives – SAA’s were the most appropriate strategy for Australian investors.
According to Naeimi, ignoring the inherently cyclical of markets by definition leads to inefficient outcomes. “Relying on long term averages to allocate capital over a five to ten-year time horizon is like having one foot in a fridge and one foot in an oven and expecting that on average it will be okay. Having a strategic asset allocation ignores the profound effect that cycles can have on returns,” Naeimi said. “The reason why dynamically managing a portfolio makes sense is because just as trees don’t grow to the sky, cycles always prevail,” he said.
For these reasons, AustralianSuper also applies a more dynamic approach to asset allocation, focusing on exceeding objectives, rather than meeting hard targets, when constructing portfolios. “Ultimately our objective is to make as much return as we can on a risk adjusted basis, so if we can deliver incremental returns without adding substantial risk over and above that objective then we try and achieve that,” Barker said.
Cooper said that while rebalancing is an implementation consideration, the real question is why have an SAA in the first place. “I think the problem we have built ourselves as an industry is that we have constructed everything based on this premise that SAA worked really well during the 1980s and the 1990s, [and that it would generate] a relatively stable return profile. The problem now is of course that one if you go back through history that is certainly not the case. There are long periods of time where that fixed allocation has not delivered the sort of return profile that people would have expected,” he said.
While Cooper also flagged dynamic asset allocation (DAA) strategies as an alternative – he pointed out that from a rebalancing perspective, these strategies face similar implementation challenges. For that reason, he said an objective-based asset allocation framework, where the investment objective – rather than the asset allocation – becomes the focus, could be more appropriate. “Certainly if you look at some of the more sovereign orientated funds, and I put NSW TCorp in this category, the more appropriate alternative I would suggest is one that is more of an objective based asset allocation framework, where the focus moves back to the investment objective you are trying to outperform or to deliver for the underlying investors,” he said.
The "Rebalancing revisited" webcast is part of the Investment Implementation Webcast Series. For further information and registration details click here.