The FHFA (Federal Housing Finance Agency) wants input on modernization in the world of housing finance.  A big change.

Human response to change is – resistance.  A natural human first reaction is anger. Change is first seen as a threat. “Who are they to disturb my nice comfy world?”  That threat of loss triggers us.  How unfair.

I commend the FHFA in the current request for information (RFI) regarding appraisal modernization.  The FHFA must deal with other, broader issues however.  It is important to recognize that appraisal is a small part of the financial world.  Appraisals are a subset of valuations, and valuations are a subset of collateral decision-making, and collateral decisions are a small part of overall risk policy management, including risk models.  It has been a while since regulations have been seriously re-examined.  It is time.

For context, remember that traditional appraisal continues to be a smaller part of this whole.  Why?  Times, they are a’changin.  But appraisal has not.  Appraisal (done by a licensed appraiser) is circumscribed by law, by standards, by legacy education, and by the demands of the very clients who would benefit.  These clients in turn, themselves, are circumscribed by various forms of regulation and organizational inertia.

We have seen the advent of AVMs, evaluations, broker opinions, waivers, and hybrids.  Each of these does better or worse depending on how you look and your motives.  Recall the three-part tradeoff of the valuation product:  speed, cost, accuracy.  “Choose any two.”

This tradeoff is the essence of today’s valuation reality:

• AVMs are cheap and quick, but do not always work.
• Appraisals are better, but cannot be reliability-measured.
• Hybrids, evaluations, and other systems separate and specialize work.

The problem with each of these and other proposed solutions is that they are coarse.  Each decision is yes/no, one way or the other way.  And each of these have a ‘basket’ of accompanying conditions, few of which optimize on cost, speed, or accuracy.  This basket effect enforces sub-optimal results.  Hybrids, evaluations, and every modified valuation type is an attempt to refine the speed-cost-accuracy tradeoff.

Is it possible that the reason for the tradeoff coarseness is because of some other reason?  Let’s look.

What lenders, regulators, and investors really need to know is risk.  Not point value.  The point value two weeks ago is just one data point toward the real question:  risk/reliability.  This not the aim of valuation.  A point value alone provides no risk nor precision nor validity information.  It is only useful when combined with other reliability measures.

Current valuation methods assume and enforce the idea that valuation delivers “market price” — mislabeled “market value.”  This “market value” was ideal in the past.  It was ideal in the days of onesey-twosey comp collecting, and pencil/paper adding machine arithmetic.  Now we do better.

Systems continue to input and respond to this single point-value number.  But there is a question:

Do lenders, regulators, and the public need market price, or something else.  A good look at this question produces the obvious answer.  It is not.

Point value is the underlying problem #1 with today’s collateral assurance modernization challenge.  Next week’s Analogue Blog asks about underlying problem #2 – the coarseness effect.