Turning risk-return theory on its head

Turning risk-return theory on its head

Business Times - 27 May 2009

Evidence from the US and Singapore markets show that high-risk stocks tend to generate lower returns and vice versa, reports GENEVIEVE CUA

FINANCIAL theory suggests that assets with higher risks should reward investors with higher returns. But that, sadly, is not the case in reality, whether one looks at the US or the Singapore market.

Concern over how retail investors are to accumulate assets for retirement, particularly in the wake of deep losses since mid-2008, has prompted one fund management firm to design an approach that it says is efficient, as well as relatively rewarding, for individuals.

DBS Asset Management believes the key is to take volatility as a starting point when it comes to screening for stocks, and not returns.

And, to eke out as much efficiency as possible - that is, the lowest possible costs - this volatility screen is used on an exchange traded index fund. Says the firm's head of management office Michael Hsien: 'Our approach is very simple. We buy and hold and take a long-only position within the Straits Times Index universe.'

So far, it has tested its 'volatility managed' (VM) approach on a number of indices with what seems to be gratifying results. VM is a more concentrated portfolio, pre-selecting stocks based on favourable risk characteristics.

Bear in mind, of course, that these are back-tested results and the past is no indication of the future. On the STI, for instance, the VMSTI generated 85 per cent greater returns and 15-20 per cent lower risks, compared to the STI alone. The results are for an 18-year period between 1991 and 2008.

A VM overlay on the Hang Seng Index had a similar outcome. Over the period, returns were higher by 185 per cent, and risks lower by 18 per cent. Ditto for using VM on the MSCI World - returns were higher and risks lower.

First, some background. DBS Asset Management is the only homegrown asset manager to launch and manage exchange traded funds here. To date, it manages two ETFs - the DBS Singapore STI and the ABF Singapore Bond Index fund which invests mainly in Singapore government securities. The latter has just been included in the CPF Investment Scheme.

The gestation of the VM approach lies in the conviction that there was more to some basic investment principles than most may think. One example is compounding. Investors are routinely taught the power of positive compounding through time, but few may have woken up to the harsh reality of negative compounding. That is, an asset which has suffered a fall in value, will need to rise by a much greater percentage than the magnitude of its fall just to get back to its original value. Hence an asset that has fallen 50 per cent will need to double to get back to par.

In terms of risk and reward, evidence from the US and Singapore markets show that high-risk stocks tend to generate lower returns and vice versa. For the Singapore market, in fact, the slope of the line which plots securities' returns per unit of risk between 1999 and 2008 is fairly flat. Its message is clear: You are not compensated for taking on risk.

'You have to be very careful when you pick stocks. The return to risk ratio is lower than people think . . . Our idea is to manage risk. Because of compounding, risk may work against you in the long run, so you have to manage the risk,' says Mr Hsien. The trick appears to be to sift out risk that is not compensated.

There are two types of risks that stocks exhibit, as the firm finds in its analysis. One is with respect to the market; the other is 'idiosyncratic' or specific to the stock. It is the latter that is expected to serve a portfolio well in terms of diversification.

'Diversification only works with idiosyncrasy. In a bull market, things are well correlated. But in a down market, they are forced to be correlated by things you had not thought about.

'We try to find ways to characterise idiosyncrasy. Of the 30 (STI) stocks, 15 are highly correlated, but they aren't idiosyncratic. It's just high beta, leverage-driven counters. The other 15 have something relatively unique about their industries to allow them to have their own behaviour. We are trying to maximise diversification.'

This takes us to the screen that the firm uses on the STI component stocks, based on nine risk factors, statistical in nature. The stocks are ranked, and the top 15 are chosen for the VM portfolio. These include SPH, SingTel, ST Engineering, F&N and ComfortDelgro.

So far, the results appear mixed. In terms of standard deviation and downside deviation, the VMSTI exhibits lower risk, but returns have suffered relative to the STI. The time period, however, is short - the current year to early May.

DBSAM, however, is confident in the VMSTI's long-term potential. It has now incorporated the VMSTI into a fund - the Horizon Singapore Equity Fund, where 30 per cent of the fund is invested according to VMSTI, and 70 per cent in the index tracker. The fund has $23.6 million in assets, and charges a 1.25 per cent management fee.

The 70 per cent allocation to the index is in line with what active managers do. 'It's impossible to be 100 per cent active, the volatility swing would be too high. Over any period of time, you'd end up with 70 per cent exposure to the benchmark . . . We're highly confident that our expenses will be lower, and investors will have a better outcome. We want to prove that and get a track record for this. We think this will work,' says Mr Hsien.

Such a strategy is of course difficult for individuals to execute on their own. But they can still benefit by having a larger allocation of their core holdings in a low-cost STI ETF, rather than an active fund that largely hugs the benchmark anyway. The ETF has a management fee of just 20 basis points a year.