Statistical Arbitrage Trading Flashcards
(12 cards)
Security market line
Expresses the excess return of an asset as a function of the market excess return.
Quantities are evaluated over a long time horizon. CAPM expresses long-run predictions for assets (securities).
How to avoid mis-pricing?
Using risk factors
Asset prices can be predicted using a small number of risk factors.
The idea is to compare assets and factors, and set prices based on such relationship, in a way that mis-pricing is avoided.
Evaluation of prices
Factor Models
Factor models evaluate asset prices from linear combinations of a small number of variables related to factors with explanatory power.
Trading factors
Factors are useful for pricing, but when factors are assets, we can trade them. The pricing model becomes a trading model for statistically hedged portfolios.
Once you know that factors drive returns, it is clear that:
* Owning assets is indirectly owning combinations of factors.
* Trading factors directly targets specific sources of return you believe will perform well.
Pairs Trading
Extends the idea of factors model beyond the pure pricing framework… a more general correlation framework.
Taking opposing positions in two historically correlated assets.
Spotting statistical arbitrages
Factor pricing models allow us to spot statistical arbitrages; whenever a market price is different from what is predicted by the model, we have a statistical arbitarge opportunity.
This is true if the model is a good model.
These are statistical arbitrages based on expectations, based on historical data.
The idea of statistical arbitrage trading
Is to use tradeable assets as factors of the pricing model and chase correlations among assets.
Trading based on statistical arbitrage
Most of speculative trading is essentially an attempt to buy undervalued assets and sell overvalued ones.
Trading based on statistical arbitrage adds more structure: by using other assets, or baskets of assets, to determine if an asset is overvalued or undervalued. Trading signals suggest to implement positions on all assets involved, both the ‘explanatory’ and ‘explained’ assets.
All trading strategies suggested by statistical arbitrage models take the form of pairs trading.
Pairs trading: CAPM
Overvalued asset
Sell the asset but also buy the market (e.g., index future).
We do not want to be exposed to market variability.
Hedge away market risk to create a market-neutral position such that any market-wide movement affects both legs equally.
Cointegration of two stocks A and B
If causality is not clear, we do not know which asset is mispriced. But we trade both assets.
If the price of A is above the theoretical price of A based on B, we sell A and also buy/sell B (depending on the sign of correlation).
Flattening Trade
Trading the yield curve
Expect the curve to flatten.
Long long-term bonds and short short-term bonds.
Reflects monetary policy expectations and inflation forecasts.
Steepening trade
Trading the yield curve
Expect a steepening yield curve.
Short long-term bonds and long short-term maturities.
Reflects monetary policy expectations and inflation forecasts.
