FOCUS 130/30 managers learn lessons of volatile markets


Date: Wednesday, February 20, 2008
Author: Claire Milhench. Thomson News.com

Quantitative managers are now reviewing their models after coming unstuck in 2007.

LONDON (Thomson IM) - 130/30 funds are learning the lessons of volatile markets, with quant managers reviewing their models after coming unstuck in 2007, whilst fundamental manager F&C says that its offering stood up to the market gyrations.

130/30 - itself a nebulous term - has met with some scepticism since launch. For a few years quant managers made all the running, with traditional fundamental managers derided as being unable to pick shorts as well as longs. But fundamental managers like F&C claim that 130/30 can work, if you have done your homework on risk and know where the skews are. The group's 130/30 fund is up by 420 basis points against the index to the end of January, since its inception in August.

'It's not just about whether you have the investment management competence to short. It is about whether you have the operational support to do it - the IT and the risk management systems,' said Peter Lees, head of UK equities at F&C. 'We have outperformed every month and all the alpha is generated on stock selection.'

Whilst generalisation is difficult in the 130/30 world, last year the quant models came a cropper, whilst the fundamental managers arguably did not suffer as much - something of an about turn considering the prior criticism traditional fundamental managers have had to take.

Quant managers got into difficulties in August when hedge funds needed to realise cash to meet their margin payments. Previously, hedge funds had taken advantage of low market volatility to leverage up in order to magnify their returns. But the credit crunch forced hedge funds to deleverage.

'The impact of that was that 130/30 managers were buying and selling the same stocks with a fairly high degree of overlap,' explained Mark Webster, senior investment manager at SSgA, one of the leading quant 130/30 managers. 'Models vary from house to house but the principles are similar, and when you get that kind of volume hitting the market quite quickly, then the share price movement can be very sharp.'

Webster said that against this backdrop, SSgA's investment signals and those of its competitors were very poor. But overselling subsequently helped stocks bounce back. 'August was like a V-shape in terms of performance for most quant managers, with a very sharp underperformance followed by a very sharp outperformance. If you had deleveraged, you'd have suffered, but had you not done anything, you would have weathered the storm okay.'

Indeed, Andrew Barber, global head of manager research at Mercer, said that 130/30 funds weren't forced to degear to same the extent as the hedge funds, so in most cases they had one negative week and then bounced back. A Morningstar survey of 38 quant 130/30 funds completed at the time supports this view, with performance roughly flat for August compared with a 1.5 pct gain for the S&P 500.

But in November, investor fears over the performance of value stocks linked to concerns about the economy, meant that signals like EBITDA, P/E, price to book, and price to cash flow performed very poorly, spelling further trouble for quant funds. 'Most managers use some form of valuation within their assessment of companies, so anybody who does it on a systematic basis like quant investors did struggle in the second half,' Webster said.

If their models cease to function properly, then quant managers face losing money on both their longs and their shorts. As a result, 2007 was a bad year for quant 130/30 overall: 'Arguably lots of managers trying to do the same thing with their models made it harder to add value,' said Barber. 'A lot of the models have valuation disciplines and being a value manager last year was not a good approach. A lot of signals became muddier. Quite often quant models will give positive signals based on underlying valuation measures and these became very contracted, so the difference between the top-ranked and the bottom-ranked stocks wasn't as great as usual. When you get very strong signals the models may work well, but when the signals are less strong, the top 50 stocks that they are buying perform less well and they are losing money on transaction costs.'

F&C's Lees argued that quant funds struggled because they were structured using back data from when market volatility was low. 'They anticipated that correlations would be very tight and that large, mid and small caps would all perform the same,' he said. 'They hadn't planned for big changes such as the credit crunch, when correlations broke down.'

F&C's approach was to short banks as they were plummeting, whilst going long oil and mining stocks. 'When correlations broke down, the number of stocks in the portfolio came down,' Lees explained. 'We took bigger positions and then shifted slowly into recovery stocks. So it was a barbell approach with some very large and very small companies, whilst we basically sold mid-caps.'

2007 may come to be viewed as a watershed year for 130/30 as this is the first time the strategy has been tested in adverse market conditions (see chart). The question now is whether institutions will still want to invest.

ViX over 12 mths
ViX over 12 mths

'It depends on the model you are looking at,' said Barber. 'Is the space they are operating in too crowded? Are there too many players trying to do the same sort of thing? And do they need to get cleverer about the way they build their quant models? I think they do have to evolve over time or the anomalies will be arbitraged away as more people become aware of it. They have to recognise that over time their edge will get competed away.'

Webster said that SSgA, which has some 12.5 bln usd under management in this strategy, was working hard at improving its models. 'We haven't made radical changes since last year because that would be the wrong thing to do. But having said that, a lot of the research that is going on at the moment into different valuation techniques shows that there are plenty of things you can do.'

One approach is to have a dynamic weighting scheme, so at certain times of the cycle, or depending on what is happening within the macro environment, weightings to value will be reigned back. 'But that's quite difficult to do,' he said.

'Another way is to have different combinations of value measures, so you use six or seven measures as opposed to two, say.' Over the long term this will dilute the effectiveness of the best valuation measure, but it will help avoid steep underperformance on a particular valuation measure.

'We are also looking at completely new valuation techniques, like implied value within companies such as residual income models or discounted cash flow methodologies,' he said.

Webster argued that quant managers had learned a few lessons from the rapid sell off in August, and the models will consequently become a lot more robust. 'We've got a strong number of ideas in terms of trying to insulate against those things and if we were to go through that period again we'd be much better equipped than we were in 2007. We don't think the model is broken because over the long term it works very well, but there are things we could have done a little bit better.'

By Claire Milhench: +44 (0) 20 7422 4808; claire.milhench@thomson.com; www.thomsonimnews.com