Serving the Quantitative Finance Community

 
User avatar
aguelmame
Topic Author
Posts: 11
Joined: December 8th, 2007, 4:10 pm

Term deposit hedge

April 25th, 2017, 8:02 am

Hi all,
In retail banking, fixed rate term deposits comes with an implicit option to the client to early withdraw his deposit with no cost.
Client option may be modeled as a bermudean call swaption indexed on rates + spread.
I would like to setup a partial hedge of this swaption. By partial, I mean a portfolio of IRS only (no credit instruments), whose aim is to minimize the hedged position P&L variance.
I was considering using the delta of a bermudean swaption whose strike is the term deposit fixed rate minus the bank's initial spread.
Does somebody has a better option or some references to articles tackling this subject ?

Thanks in advance.
 
User avatar
outrun
Posts: 4573
Joined: January 1st, 1970, 12:00 am

Re: Term deposit hedge

April 25th, 2017, 8:43 am

I would look for two things:
1) what will trigger the client to early withdraw deposits? Is it unconditional or does the money need to stay in your bank? One driver will be (better) deposit rates, but other drivers are economical. The clients might get unemployed under some scenario and in need of cash. People might start to move homes a lot more and need cash for that.
2) Is variance a good thing to minimize? What about the number of re-hedges that will costs you money every time you have to? I would look at the stability of the hedge in time. Also perhaps you want some exposures to stay within hard bounds?
 
User avatar
aguelmame
Topic Author
Posts: 11
Joined: December 8th, 2007, 4:10 pm

Re: Term deposit hedge

April 25th, 2017, 9:43 am

Hi outrun,
1) We apply statistical estimates to calibrate what I may call "structural" early withdrawal. This is estimated on periods of low rates when clients are less likely to early withdraw. So what remains to estimate is the "deposit rates" linked withdrawal. If this is correctly estimated, one could estimate a hedge ratio for the pure IR part.
2) You're right, there are other constraints to consider. I view variance minimization as a first step toward a "good" hedging strategy.
Another point I'd like to mention is the need for a risk indicator : we need to measure the sensitivity of banking book margin to interest rates moves. Regardless of whether the exposure is actually hedged, we need to report the "amount of IR risk".  My approach is to report the amount of IR risk as that of IRS portfolio that minimizes P&L variance.
 
 
User avatar
outrun
Posts: 4573
Joined: January 1st, 1970, 12:00 am

Re: Term deposit hedge

April 25th, 2017, 10:02 am

"Deposit rate linked withdrawal" is amongst other things linked to IR movements. One thing I've observed is that it matter how you got people to deposit. If you have deposits because you temporarily advertised the highest rate in the market then those deposits are "opportunistic", and those will flow out as soon as some other bank improved on your rate. There is clearly two types of customers here, sticky and non-sticky, and their ratio changes over time (the non-sticky outflow is much higher of course)

 "amount of IR risk" is that for regulatory reporting, like interest rate risk in the banking book? AFAIK those reports want you to include some element of customer behaviour tied to interest rate scenarios --i.e. withdraw rate changes when rates change-. Modelling it as an option makes sense. People won't be 100% rational, but they will be partially?
 
User avatar
aguelmame
Topic Author
Posts: 11
Joined: December 8th, 2007, 4:10 pm

Re: Term deposit hedge

April 25th, 2017, 1:23 pm

Yes, there are many reasons customers may withdraw their deposits. There is however for sure an IR driver. What I would like to do is to capture this sensitivity to IR and hedge it for sticky customers (I can't hedge against competitors boosting their rates and must simply assume this risk).

Yes, I was referring to IRRBB.
An approach like I sketched (using delta hedging) intends to capture customer behaviour sensitivity to interest rates (the delta is linked to the (risk neutral) probability of early withdrawal).
 
User avatar
DavidJN
Posts: 242
Joined: July 14th, 2002, 3:00 am

Re: Term deposit hedge

April 25th, 2017, 3:05 pm

"In retail banking, fixed rate term deposits comes with an implicit option to the client to early withdraw his deposit with no cost."


What country are you living in? Where I live cashable fixed rate term deposits pay lower interest than non-cashable deposits. Either that or one gets paid a significantly lower penalty rate of interest upon cash out. 

A very wise person once told me that if the options embedded in retail financial products were exercised rationally, retail banking would not make any money. Fortunately for retail bankers, that is not the case. As already noted, people make financial decisions for non-financial reasons.
 
User avatar
aguelmame
Topic Author
Posts: 11
Joined: December 8th, 2007, 4:10 pm

Re: Term deposit hedge

April 25th, 2017, 3:38 pm

Hi DavidJN
I live in France.
Your first remark does not imply that there's no option, it only implies that the option is initially more or less out of the money.
I don't think all retail customers will make rational decision, but a significant part will do (in fact, we consider only term deposit above a given threshold, to capture customers the most likely to act rationally).
I totally agree that people make financial decisions for non-financial reasons, but this is not true all the time.When rates rise, people withdraw their deposits, and when rates fall, they prepay their fixed rate mortgages !
 
User avatar
ppauper
Posts: 11729
Joined: November 15th, 2001, 1:29 pm

Re: Term deposit hedge

April 26th, 2017, 6:06 am

In retail banking, fixed rate term deposits comes with an implicit option to the client to early withdraw his deposit with no cost.
Client option may be modeled as a bermudean call swaption indexed on rates + spread.
question:
why would this not be a putable bond?
by modeling it as a swaption, you're assuming the money is moved from a fixed rate bond to a floating rate account at your bank?
and (you say "bermudean") this can only be done on set dates?

as an aside, many years ago I had an argument with someone who claimed to be from bermuda
finance uses the term "bermudan" while the correct adjective for things related to bermuda is "bermudean" and supposedly this is due to an error in an early microsoft spellchecker or something along those lines.
My point being that a swaption is bermudan not bermudean, because that's how everyone else in finance spells it
 
User avatar
aguelmame
Topic Author
Posts: 11
Joined: December 8th, 2007, 4:10 pm

Re: Term deposit hedge

April 26th, 2017, 7:57 am

If one ignores notional exchange, term deposit can be viewed as a cancellable receiver swap (customer receiving fixed leg), which in turn may be viewed as a vanilla swap plus bermudan swaption (forgive my incorrect spelling).
Whether the swaption's exercise is american or bermudan is not very important in this context.
 
User avatar
ppauper
Posts: 11729
Joined: November 15th, 2001, 1:29 pm

Re: Term deposit hedge

April 26th, 2017, 8:44 am

my point would be that you're making things more complicated than need be.
a bond is a simpler instrument, and simpler to price, than  a swap

another hedging option would be floorlets

but what exactly are you hedging? what's your risk?
if someone cashes in one of these bonds, they take the money and run, but what happens to the bank? what do they do? do they issue a new fixed rate bond at a higher rate ?
 
User avatar
aguelmame
Topic Author
Posts: 11
Joined: December 8th, 2007, 4:10 pm

Re: Term deposit hedge

April 26th, 2017, 10:21 am

When a customer makes a fixed rate deposit for, let's say 5Y, the bank will fund the money @ euribor and will enter a receiver 5Y swap to ensure its net margin is fixed (-ClientRate + Euribor - Euribor + Swap5Y). If the customer withdraw his cash, the bank's residual position will have directional risk (the vanilla swap, with probably a negative MtM).
 
User avatar
ppauper
Posts: 11729
Joined: November 15th, 2001, 1:29 pm

Re: Term deposit hedge

April 26th, 2017, 7:34 pm

got it, you've got to hedge the swap 
 
User avatar
Martinghoul
Posts: 188
Joined: July 18th, 2006, 5:49 am

Re: Term deposit hedge

April 27th, 2017, 10:43 am

I think banks are a bit more sophisticated than just picking a fixed 5y term to hedge.  I believe that most would have internal models that compute the effective duration of a given deposit based on a variety of inputs.
 
User avatar
aguelmame
Topic Author
Posts: 11
Joined: December 8th, 2007, 4:10 pm

Re: Term deposit hedge

April 27th, 2017, 12:54 pm

Martinghoul,
My last post intended to higlight the risk in case we don't hedge the option (static hedge).
The whole point of my question is to get helpful comments on how to dynamically hedge this exposure.
 
User avatar
Martinghoul
Posts: 188
Joined: July 18th, 2006, 5:49 am

Re: Term deposit hedge

April 28th, 2017, 10:42 am

Sure, I understand...  I imagine, given sufficient data, you should be able to come up with a model that would spit out the instantaneous delta of your deposit book, as well as maybe other exposures.  You should then be able to come up with a hedging scheme.