Is diversification dead? Many investors, notably university endowments and pension funds, have been guilty of grossly underestimating the cross correlations in the supposedly diversified assets they hold, leading to huge losses in 2008.They have been guilty of slavishly followed academic portfolio theory, while ignoring market reality. Since the mid 1990's, leading endowments like Harvard and pension funds like Calpers have diversified into alternative asset classes such as commodities and hedge funds to achieve the investment Holy Grail of higher returns at lower risk, based on historical asset relationships.  However, their very attempt to collectively broaden their investment remit began to change those relationships between assets. As these huge new investors gatecrashed relatively illiquid commodity markets in particular, they inflated a classic speculative bubble, which attracted most of the hedge funds they had invested in at arms length. They were chasingsimple out-performance of the market in the form of investment Alpha, but actually accidentally overdosed on momentum chasing Beta.

Diversification is illusory when the entire financial system is imperilled and systemic factors and 'tail risk' dominate, making the overall portfolio of many professional investors riskier than it appeared based on mining historical data.  Furthermore, there has been implicit leverage operating across many markets, particularly via hedge funds, that has been exposed painfully and pulled asset dynamics together. Rather than trying to construct spurious diversification, market timing driven by a rational macro perspective has never been more important, and is perfectly feasible with the right framework. Last September I recognized and wrote repeatedly of the divergence between credit and equity markets as hugely bearish for the latter. This February, the extreme divergence in standard deviation terms of leading equity markets from their long-term averages gave a bullish stance a very high probability of success. Below I've calculated asset correlations via leading ETFs in the last 3 months to June 30th; the larger the positive number between two assets, the worse the diversification effect and vice versa; for instance oil is 83% correlated with emerging market equities, so holding both is effectively doubling your bullish bet, whereas US small cap stocks and oil movements are only marginally related.

Asset class correlations tend to converge (ie head to 1) during times of economic and financial market distress; last year was certainly characterized by economic distress and financial market instability.    The high correlations we are currently seeing are to some extent a natural result of the steep fall and equally steep recovery in investor sentiment after last Autumn's crash, but they were evident even in the first half of 2008 before the crisis erupted. No matter how far removed a risky asset class is from the investor mainstream it is dependent on economic growth and financial stability.  In 'normal' times the underlying correlation structure reflects the dynamics of the underlying asset classes, rather than all risky assets trading off of the perceived risk of the entire system. 

For investors, there are a number of implications. Much of the time, it's appropriate to take as little system-wide risk as possible, sticking with TIPS, short duration bonds etc.  When macro conditions support it, going long risk (any asset that is vulnerable to economic or financial turmoil, including commodities)  can be achieved with no more than three asset classes, probably a mix of corporate bonds, sectoral large cap equity ETFs, and perhaps some direct commodity exposure via managed futures or an ETF is the market isn't in significant contango. The key is to make these strategies flexible enough to move across assets, sectors and into cash, if need be, aiming to achieve absolute returns and capitals preservation, like a personal hedge fund. 

The biggest decision is how much to put into risk-free assets.  Right now, near zero cash yields make these assets unattractive for any one needing to generate meaningful returns to meet intermediate and long-term goals, and we are still in a broad uptrend for equities through this Summer as incoming data continues to soothe fears and central bank liquidity washes through markets.  However, bear in mind that the assumptions used by everyone from CFPs to pension fund portfolio managers to construct their ideal asset mix have ultimately proven flawed in this bear market. Like the mantra of 'buy and hold' after decades of under-performance by equities over bonds, diversification has been revealed as another self-serving delusion of the investment industry. As the old saw goes, risk is never where you see it.