Risk Parity — Equalizing Risk

Consider the widely used 60/40 portfolio (60% stocks, 40% bonds). In terms of money, it looks reasonably diversified.

But stocks fluctuate far more than bonds. So when you measure how much the whole portfolio actually swings (its volatility), stocks are known to account for about 90% of the total risk. Various sources put it at roughly 89–90%.

In other words, 'the money is 60:40 but the risk is 90:10.' It looks balanced on the surface, but the risk is concentrated on one side. Risk parity starts precisely from this point.

Risk parity is an approach that makes 'the amount each asset contributes to the portfolio's total risk' roughly similar across assets. It doesn't split the money evenly—it splits the risk evenly.

How do you do that? You reduce the weight of assets that swing a lot, like stocks, and increase the weight of assets that swing less, like bonds. As a result, the bond weight often ends up far larger than the stock weight.

The term 'risk parity' itself is said to have been first coined by Edward Qian in a 2005 white paper. That said, the place that first ran the concept as an actual fund was Bridgewater, which introduced a strategy called 'All Weather' in 1996.

Here a problem arises. If you hold a very large amount of bonds to match their risk to stocks, the expected return itself can become too low.

So real-world risk-parity funds commonly use 'leverage' (borrowing money to scale up the size of the investment). They scale up the low-volatility bond side with borrowed money, aiming to balance risk with stocks while still boosting returns.

Supporters argue that 'if diversification is good and you rebalance frequently, leverage can actually lower risk.' But leverage magnifies losses just as much. In particular, when multiple assets fall at the same time, the size you built up to balance risk comes right back as amplified losses.

The premise of risk parity is that 'when stocks fall, bonds hold up'—that is, stocks and bonds move differently from each other. So what happens when this premise breaks?

2022 was exactly such a year. As inflation surged and rates rose rapidly, stocks and bonds fell at the same time. The correlation between U.S. Treasuries and stocks is estimated to have risen to about 0.65 that year, meaning they moved 'in the same direction.'

As a result, risk-parity strategies suffered large losses. One index bundling several hedge funds (the HFR Risk Parity 10% Volatility Index) was down about -19.5% that year, and Bridgewater's All Weather fund was reported to have lost about -22%. This is said to have been even larger than its loss during the 2008 financial crisis (about -20%). For reference, a global 60/40 benchmark was down about -16% the same year.

Risk parity offers an important insight: 'proportion of money ≠ diversification.' It makes you look at where, and how much, the risk in your portfolio is concentrated.

At the same time, its limits are clear. First, if the relationship (correlation) between assets suddenly changes, the premise breaks. Second, leverage magnifies losses, and third, borrowed money carries interest (cost). In particular, when rates rise, this cost grows and becomes a burden.

In the end, there is no 'magic that perfectly eliminates risk.' Whatever the allocation, you must look at the maximum drawdown, the loss duration, and the costs together. On 'The Return of Almost Everything,' the site I built myself, you can compare various assets using real data, and we show even the drawdowns during crisis periods without hiding them.