BPO's, CMA's, and appraisals are all used to help different parties figure out some form of "fair market value".  I think their definitions are adequately covered elsewhere -- what I am most interested in exploring in this post is their use, particularly by banks / lenders.

Before any money exchanges hands in a loan, lenders require a satisfactory appraisal on file, showing that the home's theoretical "value" is equal to or greater than the amount of the loan.

Then, before a home is placed on the market for sale, the listing agent (if they're doing their job) prepares a CMA with a list of comps to help the seller in determining a suitable list price.  This is often a very different number from the appraisal.

But, before a short sale is accepted by a lender, or a foreclosed home is returned to market as an REO sale, the lender orders a BPO (and sometimes an appraisal as well). This BPO, whether in line with the earlier appraisal and CMA or not, often determines what the bank is willing to accept as a sale price in any transaction. The BPO may be internal or a drive-by, and may be done by a local agent or one from another area -- usually not by the listing agent -- which leaves some room for variation in how close an approximation to "market value" this number actually is.

So, which is "better"? Which of the three is more accurate?

I would argue that a well put together CMA is most likely to approximate "truth" as to what a motivated buyer would be willing to pay for a particular home in a given market at a particular point in time (but then I'm biased). However, even though I think a CMA would probably do a better job of providing a lender with necessary insight, there is of course some logic in lenders not relying on CMA's from listing agents as a guide (particularly with short sales), due to our obvious vested interest in seeing the property sell.

But, the question for me is . . . why do lenders often rely on appraisers / appraisals when lending money, and agents / BPO's when collecting on the debt?

Theoretically, appraisals are more detailed and accurate. They also cost more.  So is it that lenders are just going the more cost-effective route in saving $200-$300 by settling for a BPO, or do they feel there really is little difference between the two, and that either is sufficient?

If it's that a BPO is not as accurate, just cheaper, this would seem another example of "penny wise, pound foolish" within the financial sector. Selling an REO for thousands of dollars below where it should be, based on a low BPO, or losing a deal by holding to a BPO that comes in too high, makes the couple of hundred bucks in savings laughable.

If it's that an appraisal and BPO are pretty much the same thing in terms of effectiveness and reliability, and therefore interchangeable, would a BPO not then be sufficient for originating a new loan? The lender would still have confirmation from an independent 3rd party that there's a house there, worth somewhere close to the loan amount.

In my opinion, the over-reliance on BPO's and appraisals is problematic in its own right, but this inconsistent practice of depending on one source of valuation in the loaning of money and another source in the collection of that debt -- if that's what it is -- is one I'm interested in learning more about.