The
Consumer Financial Protection Bureau (CFPB) has developed
two new disclosure forms designed to make the mortgage
process more manageable for borrowers. The first two
articles in this series showed that the new disclosures will
not protect borrowers from unjustified price changes while
the loan is in process, nor will it help borrowers shop for
the best price. This article considers whether the
disclosures will help borrowers select the type of mortgage
that best meets their needs.
The new Loan
Estimate form retains the Annual Percentage Rate (APR)
disclosure from Truth in Lending, which makes sense. The APR
is a good measure of the cost of the loan to the borrower
over the period the borrower has it. The problem with the
APR has always been that it is calculated over the full term
of the mortgage, though very few borrowers have their
mortgage for the full term. Here is an example of how the
full-term APR can lead borrowers with shorter time horizons
astray.
On December
27, 2013 a borrower with strong credit credentials shopping
my site for a $200,000 loan was quoted 4.375% on a zero-fee
30-year fixed-rate loan, and also 3.375% with upfront fees
of $19,000. Which is better? The APR on the first loan is
4.375% and on the second it is 4.19%, suggesting that the
low-rate/high-fee loan is better.
However, if
the APRs are calculated over 5 years rather than 30,
the APRs are 4.375% and 5.67%. Compressing the fees
into a shorter period raises the APR on the high fee loan.
The borrower in this case has to expect to be in the house
for more than 14 years to make the low-rate/high-fee loan
the better choice.
There are two
possible remedies for this problem. The best is to ask
borrowers to provide a best guess as to how long they will
have the mortgage, and calculate the APR over that period.
An alternative is to calculate the APR
over several periods. While borrowers would have to do their
own interpolations, this would be far better than
encouraging them to believe that the APR calculated over the
full term applied to them.
The CFPB has done neither. Only one APR
is shown on the Loan Estimate, and it is the same full-term
APR that is on the TIL.
The APR on
adjustable rate mortgages (ARMs) has another problem. An APR
calculation requires an interest rate for every month the
loan is in force. On ARMs the rate is known only for the
initial rate period. The rate that kicks in after that is
based on the value of the interest rate index at that time,
which is not known at the outset.
The assumption
used in the APR calculation for ARMs is that the interest
rate index remains unchanged through the life of the loan –
a “no-change scenario”. A 5/1 ARM that was available on
December 27 at 3.25% with zero fees had an APR of 2.98%, the
result of assuming that the index rate of 0.584% at that
time remained unchanged for 30 years.
Of course, no
assumption about interest rates over the next 30 years is
going to be right. To be useful to borrowers, the APR on
ARMs should be disclosed using alternative scenarios that
are likely to bracket the possible outcomes. A no-change
scenario could be usefully combined with a worst-case
scenario, where it is assumed that the rate on the ARM
increases by the maximum amounts allowed by contractual rate
adjustment caps and maximum rates.
Consider a
borrower trying to decide between the 30-year FRM at 4.375%
and zero fees, and a 5/1 ARM available at the same time at
3.25% and zero fees. The APR on the FRM is 4.375% regardless
of future rates. The ARM has an APR of 2.98% on a no-change
scenario, and 6.13% on a worst case scenario, which is not
very helpful. But that is because the ARM APRs are
calculated over 30 years. Here are some other worst-case
APRs on the ARM.
5 Years: 3.25%
8 Years: 4.39%
12 Years: 5.38%
30 Years: 6.13%
Note that at 8
years, the worst case APR is very close to the APR on the
FRM. This means that a borrower who expects to be out of the
house within 8 years will do better with the ARM. That is
useful information.
The upshot is
that the APR would be useful to borrowers in making mortgage
selections if it were calculated for multiple periods, and
if on ARMs it was calculated on both no-change and
worst-case scenarios.
Instead, CFPB
has decided to leave the APR as it is, and to add 3
additional measures: total interest paid over the loan term,
total payments of all types over 5 years, and total
principal payments over 5 years. These measures seem to have
been selected because borrowers understood them, but that
does not make them helpful in choosing between different
mortgage types. For that purpose they are largely useless.
