High risk should mean high returns, at least over the long run. You've got to speculate to accumulate. From this perspective, the long-run aggregate performance of Alternative Investment Market (Aim) shares is odd. Since its inception in 1997, the FTSE Aim index has lost 25 per cent - including dividends - while the main-market All-share index has doubled your money.
This isn't because small stocks generally have done badly. Aim has underperformed the FTSE small cap index by 70 percentage points since May 1997.
So, why have Aim shares done so badly?
One reason is that they carry the wrong sort of risk.
By one measure, Aim stocks are low risk. The beta on the Aim index with respect to the All-Share, based on weekly returns, has been just 0.58 since May 1997.
This matters, because beta measures unavoidable risk - the risk you cannot diversify away by holding other stocks. Economic theory tells us that it is only this risk that should generate high returns. The market doesn't reward you for taking risks you can insure against. And the trouble is that Aim stocks carry these unrewarded risks. I mean this in three senses.
To see the first, let's look at that beta more closely. In the worst week of the financial crisis (the one ending October 10 2008) the All-Share index lost 20 per cent, but Aim fell only 18 per cent. However, in the following three weeks, the All-Share recovered a little while Aim fell. This generates a low beta.
But it suggests Aim stocks carry a different risk – liquidity risk. When people panic, they sell what they can. This means that large liquid stocks do badly at times of acute distress while prices of illiquid shares might appear to hold up simply because there's no market in them. However, such stocks will drift down later as the market, in effect, re-opens and as investors rebalance their portfolios.
In this sense, assets can have low betas simply because they carry liquidity risk instead.
This distinction matters because liquidity risk is easily avoidable - you just hold larger stocks. And the market shouldn't reward you for taking such easily avoidable risks. Many investors – such as Amaranth with natural gas futures or banks with mortgage derivatives – have found that liquidity risk doesn't pay.
There's a second risk that Aim stocks carry - sentiment risk. If we look at Aim's Jensen's alpha – that portion of its returns that can't be explained by its covariance with the All-Share – we see something odd. This alpha is high around the time of bull markets in the All-Share and negative in bear markets.
This suggests that Aim shares are prone to sentiment risk. They get pushed up too far when investors are bullish, and down too far when they are bearish. Aim stocks are not traded merely on their fundamentals.
This risk isn't the same as market risk because there can be a delay between the All-Share rising and sentiment-affected stocks doing so. For example, if a bull market attracts investors' attention towards risky shares, or if it creates a climate in which people want to speculate, then sentiment shares might rise after the market generally has done so. This is what happened in early 2000. In the first two months of that year, Aim stocks rose 50 per cent even though the All-Share fell slightly. This suggests that investors were attracted to Aim as a legacy of the 1999's bull market.
Now, in theory, sentiment risk should be a priced risk, and so should be associated with good returns.
But this is only true on average over the long run. It's quite possible that sentiment risk can cause shares to fall for many years, as it takes a long time for bull market sentiment to be priced out of stocks, and because bear market sentiment can push prices down further for quite a long time. Even over a period as long as 14 years, this can cause poor average returns.
The third risk Aim shares carry is more obvious - ordinary volatility. Since May 1997 the annualised volatility of weekly returns on the Aim index has been slightly higher than that on the All-Share – 18.7 against 18.4 per cent, although of course individual Aim stocks can be very much more volatile. But this shouldn't generate good returns because idiosyncratic risk – by definition – can be diversified away.
We shouldn't, therefore, expect Aim shares to do better than the All-Share, because they carry the wrong sort of risk – little market risk but lots of non-priced risk.
However, while these wrong risks might explain why Aim should underperform the All-Share, they don't really explain why they should have actually fallen – except to the extent that sentiment risk has been adverse on average.
There are, though, two other things that can explain this.
One is a tax effect. Aim shares carry some tax advantages. Many are not subject to inheritance tax, and some qualify for Enterprise Investment Scheme tax breaks. Such advantages might look attractive. But they can produce bad pre-tax returns.
Imagine two identical shares, A and B. A capricious government then gives investors in B a tax break – say freedom from capital gains tax. Investors will then sell A and buy B. But in doing so they will push up the price of B to a level from which subsequent pre-tax returns will be poor. This process will stop when expected post-tax returns on A and B are equal.
Common sense, therefore, tells us that tax-advantaged shares will offer worse pre-tax returns than others.
Another problem with Aim has been the glamour effect. For years, investors have, on average, paid too much for 'exciting growth opportunities' with the result that growth stocks and newly-issued shares have done badly while neglected duller stocks have done well. And Aim, traditionally, has had more 'growth' stocks than dull ones.
There are, therefore, good reasons why Aim shares should have underperformed the main market. This doesn't mean they'll continue to do so; it's possible that sentiment is unusually depressed right now, and if so, Aim shares could benefit if appetite for sentiment risk returns towards normal. What it does mean, though, is that not all risks pay off, even over long periods.