馃攳
Loss reserves and capital adequacy - YouTube
Channel: Marginal Revolution University
[1]
alright the next topic is loss reserves
[3]
and capital adequacy.
[6]
We've been talking about
[7]
how we can use a mortgage pricing
[12]
model where we simulate house prices and
[17]
we look at other risk factors and
[19]
estimate the expected default costs for
[25]
a single mortgage when we're about to
[28]
buy the mortgage or ensure the mortgage.
[32]
Now we talk about loss reserves and
[35]
capital adequacy the we're going to be
[39]
using the same type of modeling
[41]
technique, but first of all we're going
[45]
to be looking at an entire portfolio of
[48]
mortgages, not just one mortgage.
[52]
Secondly, some of the loans may be
[56]
seasoned.
[57]
That is, our portfolio could include loans
[61]
that we've purchased or that we ensuring
[64]
the default risk on that we've better
[69]
have now been being paid for five years
[74]
the mortgage was taken out five years
[76]
ago. We're still responsible for the
[79]
default risk on it but we can observe
[82]
something about the path of home prices
[85]
that take took place and second and
[89]
finally we we can be focused on a
[92]
particular...
[97]
national price scenario
[103]
as opposed to looking at all of a set of
[108]
plausible scenarios and when averaging
[115]
over them we can just focus on a
[117]
particular scenario so you know it's
[121]
just like we're in the insurance
[122]
business bearing mortgage default risk
[127]
is a lot like being an insurance
[128]
business in fact there's something
[130]
called a mortgage insurance
[132]
company that does nothing but
[134]
bear the default risk prepare some of
[137]
the default risk on mortgages and
[139]
Freddie Mac and Fannie Mae are basically
[143]
in the mortgage insurance business
[145]
because they ensure they take the
[148]
default risk on loans. FHA is in the
[152]
mortgage insurance business and so on.
[157]
So, for loss reserves and capital adequacy
[161]
we would look at particular scenario for
[163]
example. So, let's... let's say we had a
[168]
seasoned loan that has gone through a
[171]
period where the cumulative price
[173]
appreciation is five percent.
[177]
So that on average low
[180]
loans issued you know let's
[184]
say several years ago have the prices
[187]
have gone up five percent, that means
[189]
that the default cost would be a lot
[192]
lower than if the loan would brand new
[198]
and there'd been no house price
[199]
appreciation in addition when you have a
[202]
season loaned you have the fact going
[204]
for you that people have been making
[206]
their mortgage payments.
[214]
One thing we can do with
[216]
seasoned loans is we could look at
[218]
the particular region or community where
[222]
the loans were originated and look at
[223]
these specific trends and home prices in
[227]
that region we can try to get a very
[229]
accurate reading of how house prices
[233]
have proceeded up until this point.
[238]
And when we're looking at a portfolio of
[240]
loans what we really and at a particular
[245]
national price scenario going forward so
[248]
let's say this is the path that's up to
[252]
now then going forward we might take a
[254]
particular home price scenario, let's say
[257]
we will go with a conservative scenario
[259]
where home prices go up on average one
[262]
percent from now on.
[264]
We still use random numbers
[267]
to say that relative to this particular
[270]
scenario where on average prices go up
[273]
one percent there's variation around
[276]
that due to regional variation that we
[278]
can't predict due to variation in
[281]
individual home prices that we can't
[282]
predict. And in fact, even for the
[286]
seasoned loans, if we have a bunch of
[288]
seasoned loans of in where the average is
[292]
up five percent we still want to put
[294]
variation around that because some of
[297]
those private those homes will have gone
[299]
up by more some who have gone up by less
[302]
and some we can declined a bit and so we
[304]
want to allow for that variation in the
[309]
historical behavior of seasoned loans so
[312]
that our seasoned portfolio won't all have
[315]
gone up by exactly five percent or
[319]
whatever they figure is we'll put some
[322]
variation in there and then going
[325]
forward we'll have similarly we'll put
[329]
variation around our average scenario so
[333]
that we get individual house prices
[337]
going up by different amounts than the
[338]
national average so we create this so
[342]
this is so we would create under a
[344]
particular scenario going forward for
[347]
average prices and we would factor in
[350]
seasoning and then we would simulate
[352]
house prices with some variation and
[355]
then estimate how many defaults we would
[358]
get and we take the total default losses
[364]
and we say that's how much we need to
[367]
reserve for losses and loss reserves
[371]
serve two functions. One is that they
[374]
actually able to create a reserve that
[376]
you can pay out you expect well on
[379]
average under this scenario that we're
[382]
reserving against, you know, let's say
[385]
we're going to have to pay out, you know,
[388]
ten million dollars in default losses
[392]
then that's when you set that aside as a
[395]
reserve that also
[397]
enters into your accounting statement as
[401]
as a deduction from income. So if next
[408]
year, house prices have done better than
[414]
you've expected in your original
[416]
scenario, you could reduce your loss
[419]
reserves and actually have an increase
[423]
in income. Conversely, if house prices
[425]
over the next year do worse than you
[428]
expect in your baseline scenario, then
[432]
you think if house prices do worse than
[435]
your income you have to take a hit to
[438]
your income. So the so loss reserves
[443]
enter is an income factor. The other
[447]
thing so that's is the first thing I was
[451]
talking about was loss reserves, and the
[455]
next thing is capital adequacy so with
[458]
capital adequacy. So with loss reserves
[461]
you the scenario you pick would be sort
[466]
of an average scenario or maybe a
[468]
slightly conservative scenario so loss
[470]
reserves you would use an average or
[475]
slightly adverse scenario. If you but you
[483]
don't do a worst-case scenario with
[486]
capital adequacy you're really trying to
[489]
do a worst-case scenario. So when I was
[495]
first doing this at Freddie Mac the
[499]
scenario was four years of ten percent
[507]
decline in house prices each year. And
[514]
then flat house prices on average for
[520]
six more years, remember when a house
[522]
prices are flat on average individual
[524]
prices will be going up and down and to
[526]
resolve that you'll be experiencing
[528]
defaults. So overall,
[531]
it's a forty percent drop in prices
[535]
spread over four years and that was what
[538]
we called the Great Depression scenario.
[541]
Now I think subsequently after I left
[544]
Freddie Mac they altered the Depression
[547]
Great Depression scenario but it was
[549]
still a pretty severe scenario. And under
[554]
that scenario if you simulate a
[556]
portfolio you'll be helped by seasoned
[559]
loans as long as there's been some
[561]
appreciation in the past but otherwise
[564]
this kind of a drastic drop at home
[566]
prices is gonna cause a lot of defaults
[569]
if in fact with this much of a drop in
[573]
home prices you're actually helped by
[576]
variation in home prices. That is the
[580]
ones that drop anywhere near this forty
[584]
percent average are all going to default
[587]
in the few that are exceptions that that
[590]
vary around this in a positive direction
[593]
only falling by ten or fifteen percent
[596]
are the ones that are only ones that are
[598]
going to not default in that kind of
[600]
scenario. So the what that gives you then
[607]
is an amount of capital that you would
[609]
need to absorb all the losses in this
[612]
kind of an adverse scenario so again you
[615]
take a worst-case scenario and use that
[619]
to arrive at capel adequacy and a point
[623]
I want to make is that all of these uses
[627]
of all these issues of pricing mortgages,
[632]
setting a loss reserve, and deciding on
[637]
how much capital you need to set aside
[639]
to survive all them involved to some
[642]
degree simulating home prices in case of
[645]
pricing models you're simulating kind of
[648]
all possible you're trying to account
[651]
for all possible scenarios for national
[653]
crisis but in case of loss reserves and
[656]
capital adequacy you could look at a
[659]
particular scenario and then vary your
[665]
vary house prices going forward around
[668]
that have the variability of individual
[671]
houses relative to that national
[674]
scenario so there's a lot of still a lot
[677]
of modeling, a lot of simulation were to
[682]
arrive at these kinds of calculations.
Most Recent Videos:
You can go back to the homepage right here: Homepage