Scribd
Upload
0 views1 page

task1

The document analyzes the adversity profile of clients over various time horizons, revealing that adversity rates generally increase as time extends, with rates ranging from 40-47% at 5 seconds to 42-62% at 30 seconds. It categorizes clients into three toxicity levels: high (Clients E and F), moderate (Clients C and D), and low (Clients A and B), with the latter two showing signs of saturation in their adversity rates after 15 seconds. Overall, the findings indicate differing client behaviors and the impact of time on trade adversity.

Uploaded by

Arkapratim Debnath
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0 views1 page

task1

The document analyzes the adversity profile of clients over various time horizons, revealing that adversity rates generally increase as time extends, with rates ranging from 40-47% at 5 seconds to 42-62% at 30 seconds. It categorizes clients into three toxicity levels: high (Clients E and F), moderate (Clients C and D), and low (Clients A and B), with the latter two showing signs of saturation in their adversity rates after 15 seconds. Overall, the findings indicate differing client behaviors and the impact of time on trade adversity.

Uploaded by

Arkapratim Debnath
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Task 1: Analyzing the Adversity Profile

How We Did It
We simulated the adversity profile for every client across time horizons τ ∈ {5, 10, 15, 20, 25, 30} seconds. For
each trade, we calculated the LP’s PnL using this formula:
PnL(τ ) = Side × Volume × (Mτ − Trade Price) (1)
We labeled a trade as adverse whenever PnL(τ ) < 0. To get the adversity rate, we calculated the percentage of
adverse trades for each client at every horizon.

Figure 1: How the adversity profile trends across time horizons τ for all clients.

What We Observed
1. Adversity Grows Over Time
For most clients, the adversity rate goes up steadily as τ increases. At τ = 5s, the rates sit between 40-47%, but
they climb to 42-62% by the time we reach τ = 30s. This makes sense because there is a higher chance of price
diffusion crossing the spread when you look over longer periods.

2. Clients Behave Differently


We saw distinct baseline levels of toxicity:

• High toxicity: Clients E and F consistently showed the highest adversity (reaching 55-62% at τ = 30s)
• Moderate toxicity: Clients C and D stayed at intermediate levels (46-52%)
• Low toxicity: Clients A and B had the lowest rates (42-46%)

3. Signs of Saturation
Clients A and B acted a bit differently. Their adversity rates flattened out after τ = 15s. In fact, Client A
showed a slight drop from 42.0% to 41.7% between τ = 20s and τ = 30s. This suggests their flow might be
mostly uninformed or mean-reverting, which contrasts with the continuous rise we saw for the toxic clients.
1

You might also like