Speaker: Ben Reeves

Portfolio managing the risky poker player

High probability of making GOOD profits RATHER than Low probability of GREAT profits

Using basics of probability such as variance and co-variance cov(x,y)

When you add a third player

2 aggressive players, bring in covariance —> to equalize bets

*if anticorrelated, then variance attributed to Mark and Beth is going to be lower, so Amin will be the one with the most attributable variance

Assets = players —> needs to diversified in risk

Size of individual bets (how position changes) > size of bankroll (gross notional value)

High asset correlation is not always desirables (players with similar playing styles)

P/L (Sharpe Ratio)

Good way to gage how you are affecting the total notional value in comparison to how that position changes (Stocks vs Bonds)

QUICK TIPS

• Types of risk algorithms - inverse vol/var, ERC
• Amount of Historical Data - the smaller your sample size, more estimation error (longer window might also lead to stale data)
  • How important is the new data