Question 1

What do you do?

Accepted Answer

We generate fast, accurate and insightful risk (from non-linear portfolios)​, saving money in 2 ways

Question 2

2 ways to save money?

Accepted Answer

• Fast, accurate risk avoids correcting intraday mis-hedges (or having to correct a non-hedge), especially on days when the curve moves are non-parallel.
 • FRTB regulations require accurate PL for lower capital usage - our solution all but guarantees PLAT can be passed

Question 3

How much does it save?

Accepted Answer

For a well-hedged market-making vol desk, savings of about 1.6m USD per annum should arise, equally split between the 2 sources of savings

Question 4

Is this a "nice to have" rather than a "need to have"?"

Accepted Answer

Most non-linear desks accept some level of inaccuracy in their books - it's just always been this way.
Traditionally there was a large overlay of linear (first order) risk so any non-linear error term is relatively small in comparison to the overall PL but with Volcker and FRTB capitalisation of these linear risks, a large overlay is now less desirable.
Currently, on the "big days" (when rates move, curves steepen / flatten, the vol surface moves massively and skew prices jump), Risk and PL come in far from the prediction.

Risk management want to know why but it's an unfair question to the desk - they haven't been given the tools (or risk) to give a good answer.
 
But the question that risk management should be allowed to ask is:  
Can you predict the risk and PL for the next day if the markets moved exactly the same way again tomorrow (or if it moved 200% or 300% of today's move). Traditionally this would only be possible by running those exact scenarios - with PLATSON, the answer can be given in seconds and adjusted pretty much ad nauseum.  This insight suggests what the natural hedges for the book might be enabling correct hedging of the book before the large moves (and losses) occur, rather than just buying some 1m10y straddles and hoping that helps a bit.

Question 5

When do we need it?

Accepted Answer

The hedging savings could start to come within a month of installation - the savings are immediate and the cost is only based on what is actually saved.

FRTB savings will only apply from 01Jan2022 but the solution gives the PL accuracy for free until then

Question 6

Why not just reval the portfolio and get the new risks?

Accepted Answer

1. Risk Insight
Our method can show the drivers of the delta, vega and gamma changes and thus hedges naturally present themselves. A simple reval just gives the changes and PL
2. Timeliness
Our method can be run with a simple Shift F9 on a spreadsheet.  A full portfolio reval with new delta, vega, smile parameters etc can take 20 mins to upload the new vol surface, curves, run the delta, vega bumps and then you are back to where you started with incompletely specified risks. 
Our method puts you ahead of the curve rather than behind

Question 7

How does it work?

Accepted Answer

PLATSON allows some small error terms to exist and then minimises those error terms. Traditional methods allow no error but the assumption (of a parallel curve move) induces a massive error in itself.
Traditional methods hit their target with 100% accuracy but the target is in the wrong place by about 35-65%.
The PLATSON target is in the right place but we are OK to miss it by 1-2%.

PLATSON SOLUTIONS

Higher Order Derivatives Analytics