Evaluating Skew PnL Attribution

We've discussed the benefits ofPricing Skew & Flow Imbalance that can help make your business more profitable.

There are a few means of assessing skew and which is best depends on the application.

Spread Based Assessment

This is evaluating against a model price against a benchmark price and determining which is making more money.

Definitions

Benchmark rate - this is flexible but defaults to be the aggregated pool market data.

Pricing model - mahi's pricing model.

PRICINGBENCHMARKMIDPNL / MidPnlDiff The difference in spread PnL from filling against Mahi's Pricing model's mid rate and the Benchmark mid rate. This is a spread neutral evaluation.

Methodology
On receipt of each trade we evaluate
- 1) if our Pricing Model was lower than the benchmark rate:

a) if the trade was a BUY for the house, then we award the difference in PnL filling via our mid vs the benchmark. We are buying for cheaper so we get the positive difference in PnL.

Example
Benchmark EURUSD 1.1

Pricing model EURUSD 1.0

We BUY at 1.0 so for a 1K order that is in counter terms:

= 0.1 * contract size

= +$100

b) if the trade was a SELL for the house we have sold for less. That is bad, so we attribute -$100 instead.

- 2) If our Pricing Model was higher than the benchmark rate

a) if the trade was a BUY for house, then we have bought for more. That is bad, so we attribute -$100 using the example above.

b) if the trade was a SELL for the house, then we have sold for more. This is good, so we attribute +$100 using the example above.

Dashboard Statistics

Risk Reporting publishes these running totals since it has started to the Trading Dashboard in the Skew Panel.

We are able to run these measures as a simulation

PREDICTIVITY BASE ASSESSMENT

Looks at how that skew turned out in 30/60/180 seconds. This is good for attracting good flow and defending against bad flow. The impact isn't necessarily shown in the spread, as it emerges over time.