Tip:
Highlight text to annotate it
X
I am Howard Lothrop host of Echo Partners TV. Lately I've been on this tear, where I
have been questioning some of the conventional wisdom assumptions in the industry. I'm going
to continue down that path today, as I look at another question that that should be asked
by users of asset liability models and interest rate risk measurement models. That question
is simply this: Do more complex instruments specific models outperform simpler call report
based models? You know what? On the surface this seems like a no-brainer. Anytime you
have more bank specific information it's an easy jump to the conclusion that says we should
have a better more accurate result. Unfortunately, I don't think that the results are all that
clear-cut. I can give you several reasons for this but I think the most powerful reason
is to me the simplest and that is by going to a more complex model with line item and
instrument level detail to really get the value of that model you've got to go to instruments
specific assumptions as well. Otherwise, if you're aggregating your assumptions you are
really migrating backwards from instruments specific towards an aggregation model such
as a call report model. And I'll go one step further on this and say that you know there
are a lot of really smart people that try to keep up with the asset liability model
assumptions, particularly with respect to securities prepayment rates or loan prepayment
speeds and you know, it's kind of like predicting interest rates. I just don't believe people
can do it successfully over the long run. The key here really is to get the direction
and the rough order of magnitude correct. So if you know that rates fall and your prepays
speed up and rates rise and they slowdown, that's half the battle. 90% of your answers
in asset liability management are going to be given to you, just by getting the specific
relationships right in a very general sense. Now I realize this can be a difficult item
to accept and so I am going to disclaim right now that you know I'm talking to community
banks here. I'm not talking to $1 billion plus institutions that have complex balance
sheets. I'm really talking to the meat and potatoes community banker that has a pretty
straightforward balance sheet. It doesn't have a lot of inverse floaters or other not
normal kinds of instruments. And since I'm way out on a limb now by saying I'm not so
sure how the performance works, I'll just sawing on the limb altogether and I'll just
throw out a challenge to you. I run an interest rate and internal proprietary interest rate
measurement model, asset liability model, on all 7200 banks nationwide every quarter.
I will challenge you if you are interested and want to see how your model stands up.
Let me know and we'll just match up our assumptions and I'll prove to you the key point here,
which is that the basic calculation and mathematical engine these models is pretty much the same.
Simulating interest rate movements is not that difficult a task. At least not with respect
to getting the most important set of data, and that's about the 80% of data that you
need to identify whether or not you have a significant interest rate risk problem . So
if you would like to take me up on the asset liability challenge, by all means, you know
send me an email or put a comment here and we'll see if we can't get together and make
a believer out of you. So thanks so much, and we look for seeing you on the next episode.