Equities' Jekyll and Hyde
The Dutch asset management house Robeco last year announced that it was launching its first value fund, benchmarked to the MSCI World Value Index and investing solely in undervalued securities.
The significance of this is that Robeco has a strong tradition of growth investing. Its Rolinco fund is a pure growth fund, and its Robeco fund has also traditionally emphasised growth stocks.
Robeco says it has introduced value investing to be able to respond to different economic scenarios. “Value funds have a lower risk profile than growth funds. Since growth investments concern investment in companies which are expected to show considerable growth, their share prices are more sensitive to economic news.
“In times of high economic growth, the growth style can be expected to generate higher returns than the value style. This is not the case when the economy slows.”
The move is a sign, if one were needed, that leading European asset managers are recognising the importance of style in their investment processes.
Style investing has developed over past 60 years, chiefly in the US where style was early recognised as a useful methodology for analysing portfolios. The main style categories that have been identified – again, in the US – are value and growth, large and small.
Value usually refers to securities with high book value to market price, and a high dividend yield. Growth is seen as the reverse of value – that is, low value stocks with growth potential.
This simple reverse of the coin, however, has become less acceptable to investors recently, particularly outside the US, who think a more sophisticated definition is needed. This is reflected in the international index providers’ approach to incorporating growth and value in their indices
Some index providers simply use a single factor – low price to book – to distinguish growth from value stocks. Standard & Poor’s indices, for example, use a price-to-book value calculation to differentiate between fast growing companies and slower growing or undervalued companies. If the price-to-book value is low they are categorised value. If the price-to-book value is high, they are categorised as growth.
However, a number of index providers have introduced more complex methodologies. MSCI, for example, switched in early 2003 from a single factor to a multiple factor methodology for its MSCI Global Value and Growth Indices. This featured a number of innovations, including the use of eight different variables – three for value and five for growth – to reflect value and growth styles more accurately.
Other vendors, such as Russell and Dow Jones, also use a multi-factor model, which combine numerous valuation techniques to distinguish between value and growth.
The Association for Investment Management and Research (AIMR) has not categorised styles in the traditional way and has instead introduced more theoretical definitions. Robert Schwob, founder and chief executive of Style Research, a UK based global equity research provider, thinks this was a mistake.
“There is something very basic about the distinction between value and growth. It has to do with the basic differences between fear and greed and also between the practices of contrarian investing versus trend following,” he said in a presentation to Ireland’s Society of Investment Analysts earlier this year.
“While we should applaud the AIMR for not wanting to lay down the US example as the only template for other markets, its theoretical definition suffers from being too general and so not capturing the important common features of investment analysis and investor motivation that lurk behind the simple definition of style.”
Schwob says the performance of the main styles has followed similar patterns across most major markets and regions. Large value outperformed during the early part of the past 20 years. Growth, mostly large growth, outperformed during the late 1990s. And small value has outperformed dramatically over the past four years.
Furthermore, he says, it is possible to forecast how these major styles are likely to perform in the future. Style Research has modelled the forecasting of style returns in terms of three cycles – the equity market cycle, the economic cycle, and the interest rate cycle.
This model shows that value performs better at the top of the equity market cycle, when share prices begin to fall, since over-promoted growth stocks will fall faster. They will also perform better at the bottom of the cycle, since defined benefit pension plans re-entering the market will buy up high-yielding value stocks.
Value also does well at both the top and bottom of the economic cycle. Worries about recession will encourage investors to invest in value stocks, while confidence about the up-turn will also encourage them to invest in smaller, riskier value stocks.
Finally, value benefits from the interest rate cycle, since investors will invest in value stocks during periods of high, longer-term interest rates. “High long-bond yields generally occur against a background of high and threatening inflation,” says Schwob. “This shortens investors’ assessment horizons and focuses them on the more immediate returns from value stocks.”
There is a consensus among academic that value-style investing produces superior returns to growth-style investing. However, there is far less consensus about why this should be so. Kenneth French and Eugene Fama, who demonstrated the value effect in 1992, started with the assumption that markets are efficient and therefore attributed the higher returns of value strategies to their increased risk.
They argued that stocks with high ratios of book equity to market value are more likely to run into financial trouble and are therefore riskier. However, critics of this approach say it is obviously flawed since it regards bubble internet stocks, with virtually no book value, as less risky than solid utilities, which have high book values.
Josef Lakonishok, Andrei Schleifer and Robert Vishny, who also demonstrated the value effect two years after French and Fama, have a different explanation. They argue that risk does not explain the differences in returns. They look instead at investor behaviour. They suggest that investors create a favourable sentiment for growth stocks by extrapolating their past performance too far into the future. As the market gravitates towards growth stocks, value stocks become under-priced relative to their fundamentals.
Lakonishok and Louis Chan, reviewing the value-growth debate in the Financial Analysts Journal, suggest that expectation errors are at least part of the reason for superior returns on value stocks. “Investors have exaggerated hopes about growth stocks and end up being disappointed when future performance falls short of the unrealistically high hurdle. By the same token they are unduly pessimistic about value stocks and end up being pleasantly surprised.”
Another area of uncertainty is whether the value effect operates as strongly in markets outside the US. Return premiums for value stocks have been documented around the world. Yet, until recently, the lack of data has limited studies of non-US markets to samples of large-cap stocks over recent time periods.
However, in 2003 Elroy Dimson, Stefan Nagel, and Garrett Quigley showed that the value effect has been as strong in the UK as in the US, and over a long period of time. Using accounting information merged with share price data they found a strong value premium in the UK for the period 1955 to 2001 in both small-cap and large-cap stocks.
Yet they warn against projecting these findings into the future. “Value investing was a winning strategy in the period covered by the major research studies but the subsequent outcome was, for several years the opposite of what history had led investors to expect.”
In particular the 1990s was a decade of growth stocks, when value strategies did poorly. However, value strategies triumphed after the tech bubble burst in March 2000 and have continued to outperform growth.
The picture is strikingly similar across the world markets, as ABN Amro’s Global Investment Returns Yearbook (GIRY) 2005 shows. The GIRY authors, Elroy Dimson, Paul Marsh and Mike Staunton, all of the London Business School, have calculated the value-growth premia – defined as the geometric difference between the return on value and growth stocks – in 17 countries during 2004.
This shows that growth stocks outperformed only in Spain and Ireland, where style effects are weak anyway. There was a positive but small value premium in Denmark and Sweden, while value outperformed growth in the other 12 countries, with premia ranging from 4% in Switzerland to 26% in South Africa.
The performance of value strategies in 2004 continues a five-year run of outperformance in world markets. Five-year annualised value premia since 2000 range from minus 1% in Switzerland – the only negative performance – to 40% in Sweden. The value premium was 13% in the UK, 14% in Germany and 10% within the MSCI World Value index.
This performance could simply reflect a dismal period for growth stocks rather than outperformance by value stocks. Yet the evidence is against this. Value strategies worldwide did well in absolute terms.
Between 2000 and 2004 only value stocks in Switzerland, the Netherlands and Germany underperformed, with annualised returns of between minus 4% and minus 5%. In the US, Japan, France and Belgium, returns were close to zero and in the remaining 10 countries value stocks provided positive returns ranging form 3% top 13%. This was against a MSCI World Value Index return of 0.4%.
Dimson, Marsh and Staunton conclude that “value stocks have delivered positive returns over the last five years, which is remarkable given the savage bear market for 2000 to 2002”.
The size effect, visible in US stocks, is also evident in the UK. This is demonstrated by the performance
of the Hoare Govett Smaller Companies (HGSC) index, which comprises the lowest tenth by value of the UK equity market, and now has a 50-year history.
It shows that £1 invested in the UK equity market at the start of 1955 would, with dividends reinvested, have grown to £549 by the end of 2004, a return of 13.4% per annum. An equivalent investment in the HGSC would have generated £1,821 or over three times as much equivalent to a return of 16.2% per year. But an investment in micro-caps would have yielded £9,834, more than five times as much as an investment in the HGSC and equivalent to an annual return of 20.2%.
Clearly, over the long term there has been a substantial size premium in the UK, with the smallest, micro-cap stocks performing best of all.
Again this effect is replicated in world markets. The GIRY 2005 looked at the small cap premium – defined as the geometric difference between return on small and large cap stocks – in 16 countries during 2004.
Small caps underperformed in only three countries – Belgium, Canada and Ireland by around 5%. In three other countries – France, Italy and Sweden – the premium was around zero. But in the remaining 10 countries the premium was positive ranging from 4% to 19%. with 7% in the UK, 11% in Germany, and 14% in the US.
Again most of the annualised returns on small-cap indices over 2000-2004 were positive. Most countries – 13 out of 16 – had positive annualised returns, ranging from 2% in Belgium to 20% in the US. The average return for the 16 countries was 7.7%.
Dimson, Marsh and Staunton observe that “for small cap investors in most countries, it is as though there was never a bear market at all”.
However, the small-cap effect has not paid of as handsomely as it should, since soon after its discovery the US size premium went into reverse. This was repeated in the UK and almost all the markets around the world. It is only since 1999 that the size effect has reasserted itself.
If style investing does not always pay, does it have a future? Or will it be a fascinating but expensive toy that only the largest institutional investors can afford to play with?
Schwob of Style Research suggests that there are a number of factors that are now combining to ensure that style will continue to gain in popularity in most world markets, in spite of the uncertainties.
The increasing globalisation of markets is forcing managers to understand and invest in foreign markets, he says. “Style represents a quick-start introduction into the patterns of investment where managers may not yet have deep familiarity and insights.”
Better still, style investing gives investment managers something positive to tell their clients. For if one style does not perform, another one will.