In the article last week we looked generally at the concept of diversification. What we want to address this week is some of the more specific considerations when trying to diversify. These are concepts you will hear talked about, and we will touch on their strengths and weaknesses in relation to diversification.
Correlation
When talking about diversification you will hear people talk about investment and the correlation, or relatedness, of the different assets held. The theory is that uncorrelated or negatively correlated assets improve diversification benefits. Correlation is simply the degree to which a movement in one asset is likely to be associated with a movement in another asset. So for example, if two assets have a correlation of 1 (perfectly correlated) a dollar gain in one asset will be associated with a dollar gain in the other. Similarly a loss will be perfectly matched.
This doesn't mean that one asset is causing the movement in the other. It would be a mistake to infer a causal relationship between the two, simply because of their correlation. One asset doesn't necessarily cause any movement in the other, they simply move in a related way. For example, BHP doesn't cause the movement in the price of RIO however they do have a tendency to move in related ways (although certainly not perfectly). Commodity prices are more likely to have a causal relationship with both companies and explain at least some of the correlation between them.
The argument for uncorrelated assets is reasonably strong. If you have a portfolio of uncorrelated assets which perform well, it would mean that strong returns probably indicates investment in superior assets rather than simply riding the back of a broad market trend. However you need to be careful with negatively correlated assets. This is because if a strong negative correlation exists, as one asset value moves downward, the negatively correlated asset's value would move upward. Clearly you don't really want high negative correlations, or you simply guarantee little or no return in all market conditions.
For diversification, it is generally accepted that to achieve a well diversified portfolio you want to hold a basket of uncorrelated assets (a correlation coefficient of less than plus or minus 0.3 for example).
Liquidity
It is also common to hear that illiquid assets are adding diversification benefits. That is, that the movement in value of these assets are not related to the movements in value of the other assets you hold. The problem with this kind of thinking is that it's quite possible that this belief if is more a case of perception than a real benefit.
The perception of low correlation is aided by the fact that many illiquid assets are not re-valued as often. This allows for valuations to span business cycles, leading to a perception of stability which is not actually there and you perhaps don't get a true indication of volatility in prices. This doesn't mean that there aren't advantages to this because it helps investors view long term assets as long term. For example, if you own an investment property you don't run out and ask someone every minute of every day what they would pay you for the property. In effect however, this is exactly what is done with shares on the ASX. This leads us to watch the evening news for the days stock movements, whether your intention is to sell or not. What it does mean is that the apparent lower correlation with liquid assets is really due to a timing effect, rather than any real diversification benefits.
Timing Effect
The problem with ignoring asset movements in other asset classes is that many of the same factors that affect more liquid markets will affect illiquid asset prices. For example 'collectibles' are often affected by a buyers perceived wealth, which is often affected by how liquid assets are fairing. As the GFC hit the share market we heard talk about how good art was as a diversification tool. However, as people lost significant wealth in stocks, the art auction clearances started to tumble and then the prices started to fall. It could be argued that Art offered diversification, but only to the extent that you got out before the timing effect diminished.
Similarly, property prices often lag changes in the business cycle due to the relative inefficiency of the property market. However, as confidence reduces, investors require a higher return to justify a purchase. In property this usually takes the form of increased yield. In poor business environments it is not often possible to increase rents (in fact they often fall as vacancies increase and incentives are needed to get new tenants) and so valuations must fall. So again, the argument for asset class diversification may be more effective for short term, less traumatic events.
Strength of Diversification
Diversification is likely to be effective during large systemic events, only in so far as you are diversified across asset risk (for example, growth investments such as shares and property versus defensive investments such as cash and some government bonds). The challenge here is that during good times when perception of risk is low, holding the more conservative risk investments can seem like an unreasonable drag on your portfolio. It takes a great deal of investor fortitude to stick to your diversification plans when for example, cash is paying 4% and shares have put on 20%pa for the last 5 years. However it is often at times like these that the value of diversification, if not apparent right now, will be in the not too distant future.
Diversification generally is a good idea. We can't get all our investment decisions correct and spreading the risk is a good strategy for reducing the chance of catastrophic loss. However, like all good things moderation is the name of the game. Otherwise your portfolio becomes unwieldy and you risk losing site of what's inside your basket.
