OUR APPROACH
Our approach to investing is based on breaking the convention of basing every investment deciision on outdated models that use incorrect assumptions to achieve closed-formula yet incorrect data.
​
We believe that most attempts at predicting stochastic processes are inherently rooted in wrong assumptions about the Gaussian distribution of returns.
​
Numerous research has proven that the standard deviation of returns exhibits much more prdictable properties, due to multiple factors (e.g. Strong mean reversion properties, provable by Phillips-Perron Test, among other methods).
​
We do acknowledge however, that any attempts at prdiction of stochastic processes carries directional risk. For that reason, at AWP we emphasize relative value, delta neutral trades. Which constitute 50% of our portfolio.
STRATEGIES AND EXECUTION
Gather variance risk premium, which we have found to be widely present across all markets and asset classes.
​
It has been widely studied and proven that option implied volatility more often than not exceeds realized volatility of the same underlying asset. This fact holds true across a variety of asset classes.
​
We seek to exploit this phenomenon across a variety of markets, primarily:
- Cryptocurrencies futures, options
- Commodities futures, options
- Equity indices
EXAMPLES
Dispersion Trades
Long / short pure volatility of an index via variance swaps. Simultaneously enter an opposite side volatility trade of selected components of the same index. Either by variance swaps, or by dynamically replicating the swap by delta hedged opposite positions.
​
Implied Volatility Arbitrage
Our study finds thats the highest premia can be extracted by selling overpriced volatility. Although not risk free, this form of arbitrage can provide significant returns, especially when the prevailing models are regularly recalibrated, and new innovations are introduced to the models based on new findings.
Although EWMA, GARCH(1,1) in combination with proprietary innovations data (interest rate, credit market, corporate actions) has proven itself as quite apprpriate for most markets, TGARCH, NGARCH, EGARCH, IGARCH, EVT, Cornish-Fisher models are also used when appropriate, with monte-carlo techniques deployed for volatility forecasting.