The subject of whether the lender knows or not whether the product is "good" or "bad" is a more nuanced assessment. They generally don't care. It's cheap and it satisfies the bank examiners. That is all they care about.
I agree with this however...
I doubt that many lenders rely upon ANY appraisal in their lending decisions.
They do for the banks I work with. Rejected two in the last 30 days. Usually, through the review process, the appraisal can raised to a level of acceptability. Not always, though.
Especially after they realized in a downturn like 2008, that the past appraisal meant nothing. The value was not there even with the best appraisers. After all, when prices fall by 30% plus, no one's appraisal was going to support this new value during the origination appraisal of the past years. So it did not prevent the collapse. it did not predict the collapse and especially could not foretell the actual price the bank would get out of their REOs. Whether someone gutted it out and paid off the loan regardless they were underwater was a testament to their integrity, the depth of their pockets, the stability of their job, and their credit history.
In a housing downturn, the riskiest loans were made to people whose livelihood originated with the housing industry itself. So carpenters, developers, cabinet installers, even brokers and agents, were the ones whose incomes fell dramatically. The local government employee, school teacher, etc. could survive the downturn much better than most. And did.
I suppose that we need to recognize that the bank appraiser is only telling the lender that they are not upside down in the loans they make before they start. And that may predict whether or not someone has enough skin in the game to be a reliable borrower. it is easy to walk away when you have no skin in the game. So the value of our appraisals may only be good on that very day we say it is and not much further out in time. Thus the banker is far more likely to rely upon ones credit and work history rather than the appraisal.
(my bold)
This is it in a nutshell.
Although commercial reports may go a little deeper in terms of supply/demand (which implies some future demand forecasts along with feasibility), residential appraisals... especially GSE-type, rarely go beyond historical (inferred on a very basic level). And, for the majority of the assignments, that's good enough and meets both the client's expectations and work done by a peer.
But the bottom line is the appraisal is only good to mark-to-market the value of the property on that date. It may provide some indication of past trends (rising, stable, declining markets) and it may go an extra mile by discussion pipeline and absorption. But for a routine residential transaction: What's the value so I can calculate the LTV and determine the loan amount. I'm not counting on getting paid back by the collateral, I'm counting on getting paid back by the borrower's ability to service the debt; that is best modeled by credit score and employment history.
That's why, not knowing what the parameters are going to be for these hybrids
if they roll out (and I think they will) as a substitute for a traditional appraisal, the question is, how collateral-value sensitive they will be?
Let's assume that a percentage of appraisers doing these (by intent or incompetence) over-value a property on a routine basis by 10%.
If the credit score and employment history is solid, and the LTV for this type of loan is maxed out at 65%, how critical is that 10% over-valuation (mind you, I'm not excusing it or justifying it, I'm just asking the question):
$300k value vs. $330k.
$195k loan vs. $214,500 loan (@65% LTV).
P&I monthly payment difference at 5%, 30-years, fixed?
$1,047/month vs. $1,152/month. $105/month for a borrower with good credit score and solid employment history.
If that borrower would qualify for the same payment at a 65% LTV (in other words, if the borrower could qualify for the home if the loan was $214,500) how risky is using the hybrid, even when assuming it is off by 10% on the collateral value?
I would argue that it doesn't add much risk at all... excepting in the case where we have a significant market downturn, and then all bets are off anyway.
If you take that 10% over-valuation, and say 50% of all hybrid transactions appraised are going to be overvalued by 10%, then the payment difference (using the values and terms I outlined as an average) looks like this:
$1,047/month vs. $1,099/month. A difference of $52/month.
Certainly, if the borrower would qualify at the 10% over-valuation, and we assume that 50% of these transactions are over-valued to a potential of 10%, then the on-average 5% overvaluation isn't that significant in terms of the borrower's ability (based on credit and employment) to service that debt.
I think these are the kinds of macro-analyses that drive much of the decision-maker's evaluations of these products.
We are focused on the per-assignment valuation (me) and the larger impact on the policy of using such products (you). It isn't that those concerns aren't part of the lender/GSE's evaluation metrics. It is that they don't just stop at the value-component; they take that to the larger component of risk mitigation.
Lastly, I hope (probably in vain) that I won't be misquoted as saying I don't care about appraisal credibility or that it is perfectly alright to overvalue properties because the laws of large numbers, or the specific inputs I use, seem to argue that it doesn't matter. That's not what I am saying at all.
I am saying that in order to put limits on these types of products, it would be in our best interests to consider how the decision-makers who will determine their use and their breadth evaluate their risks.
