This is an excerpt from Ben Inker's commentary in GMO's Quarterly Letter for the first quarter of 2017. You can read the full version here
Investors spend a lot of time worrying about what can go wrong for their portfolios. This is a worthwhile and, indeed, essential exercise, but the obvious corollary activity, protecting a portfolio from what can go wrong, is less clearly a good idea. This seeming paradox can be disentangled by recognising that it is impossible to determine if you are taking an appropriate amount of risk without understanding what the downside is for your portfolio, which means you simply have to do the exercise of understanding what can go wrong. Buying insurance for your portfolio to reduce that downside, however, is a much dicier proposition, leaving you generally either paying so much for broad insurance that you might as well own less in risky assets in the first place or buying insurance for such a narrow range of events that you haven’t protected your portfolio that much at all. This is not universally true, and it can make sense to hedge a piece of the risk that goes along with an asset whose other characteristics you like. It can also make sense to shift the assets you own in the first place given the particular risks that seem most important for your portfolio. For example, in our Benchmark- Free Allocation Strategy we are currently hedging a piece of the risk in emerging markets to protect against the possibility of US dollar strength, and we are choosing to hold alternative strategies and in inflation-linked bonds versus equities and conventional bonds to protect against the risk of rising inflation and rising discount rates.