Portfolio construction
Risk parity (inverse volatility)
Each ticker gets a weight inversely proportional to its volatility, so the calm ones get more money and the wild ones less.
The idea in plain words
Equal money is not equal risk: 25% in a volatile index carries far more risk than 25% in a calm currency. Risk parity equalises the risk instead of the money, in its simplest form by dividing by volatility. It needs neither expected returns nor correlations.
How it is calculated
Weight each ticker by 1 ÷ its volatility and normalise so the weights add up to 1 (always long only).
wᵢ = (1 ÷ σᵢ) ÷ Σⱼ(1 ÷ σⱼ)How to use it
- As a robust, simple default: it is hard to get badly wrong.
- Combine with Volatility targeting to set the overall risk level you want.
- Use ERC instead when some tickers are strongly correlated and you want that taken into account.
Drawing conclusions
- Risk shares close to equal mean the correlations are similar across pairs; very unequal shares mean ERC would differ.
- A low expected return is common: it puts the most money in the calmest tickers.
Example
Running example (used in every portfolio guide): US500, GER40, gold (XAUUSD) and EURUSD, two years of daily data. Expected returns 8%, 7%, 7% and 1% a year; volatilities 18%, 20%, 15% and 7%. US500 and GER40 move together (correlation about 0.7); gold hardly moves with the indices (about 0.04). Weights 18% US500, 16% GER40, 21% gold, 45% EURUSD. Return 4.5%, volatility 8.5%. Risk shares 27%, 28%, 20% and 25%: close to equal, but the two indices together carry 55% of the risk because they move together, which inverse volatility ignores.
The example's inputs are illustrative; the weights and figures are what our optimiser returns for them.
Watch out for
- Ignores correlation: two near-identical tickers each get a full share.
- Often under-invested in return; the institutional versions use leverage to compensate.
A little history
Bridgewater's All Weather fund (1996) made the idea famous; Edward Qian of PanAgora coined the term "risk parity" in 2005. After 2008, when such portfolios held up well, it became a major institutional strategy.
Educational content only. No portfolio model knows the future: every one of them works on estimates from the past, and nothing here is investment advice.