I'm merely saying -- and I'll say it loud and long -- that mark-to-market is NOT a reasonable and/or acceptable way to value assets in a considerable number of business activities. That there is a crying need for an acceptable evaluation method in these industries is beyond arguing.
Attempting to mark ''to market'' the value of, say, a strip mall in a depressed area, on an immediate basis, is futile and self-deceiving and an exercise in time-wasting. Like as not, in any given day and/or week, there will be no bids for such a property or any comparable property. Therefore, mark-to-market is ludicrous, no more, no less.
To value such a property correctly (or, at least, as correctly as can be done), one is going to have to take a view over time; i.e. all the bids and transactions on comparable (or nearly so) properties within the area over a year or two or three, and concoct a sort of rolling moving average.
Now, this doesn't cure all ills, by any means. The resultant valuation is still going to be a rough one, but it's going to have at its base the real-world bids and transaction values. And, it's going to be a hell of a lot more accurate than an immediate mark-to-market valuation.
This fantasy that one can ''mark to market'' something that does not have a regular and freely traded market to begin with is so wrong-headed and so corrosive to proper conduct of business that it must be junked straightaway.
To answer your question directly: were I a lender or investor in, say, commercial real estate, would I prefer immediate mark-to-market or historical bid/offer data as a metric for the most accurate possible valuation of the real estate in question, I should vastly prefer -- without any hesitation -- the historical transaction record, assuming only that the data is sufficiently recent (say, w/in the past 6-12 months or closer).
Immediate mark-to-market is BS in these markets, and this is easily demonstrated. First, nothing worth the description as an ''asset'' can ever have a valuation of zero in normal circumstances (I exclude here things like brownfield properties, which, if their real-world value is in fact zero, this is due to exogenous circumstances such as lawsuits). Yet, we read every week about some company having had assets of one type or another marked at zero.
Second, pick a property, any property, in some state. Call every potential investor/lender in the state and ask ''What would you bid for this property if you were interested in it?'' Guess what? You won't get any data, because they won't tell you. Where's your ''market'' that you're ''marking to'', eh? No bids, no market. Pretty simple.
So, what's the mark-to-market value of the property? Zero? Can't be. The property is certainly worth something. What some accounting boffin says it is? Gag me with a shovel, puh-lease! What the goobermint says it is? That's worse still; gag me with a backhoe.
Like it or not (and I don't), in such situations as these, all you've got is approximation via historical transactions.
Write this down in big, black letters: Mark-to-market is meaningless or worse when there is no stated bid/offer market for the goods or property under discussion.
Assume, for example, that a bank has a portfolio of loans that trade at a discount from nominal value. That being so, the value of those assets should be marked down on the bank's books. The bank then has to add to loan loss reserves, raise more capital, or be sold to a stronger bank.
Of course, when the general level of asset values in the market falls in a recession, there is a lot of pain to go around. But pretend accounting values for banks and financial companies does not eliminate the pain, it just disguises it and draws in new victims.
In the current crisis, the federal government is injecting capital, cheap credit, and new money into the banking system in order to prevent a freezeup of credit markets. Even so, the banks themselves do not yet trust their own or anyone else's balance sheets because, with business conditions still weakening, the markets have not yet cleared and found bottom.