Mark to Market or Not to Mark to Market?

The latest issue that has emerged in the talks on the bailout is whether the practice of marking to market, required of financial service institutions under FASB rules, should be suspended or even ended. Financial service firms currently are required to revalue the securities they hold as assets on their books at market value each period. As the markets for many mortgage-backed securities have dried up, their values have plummeted, which in turn have put the limited capital that banks and investment banks at risk.

I have mixed feelings about the rule. I am a believer that investors should be provided with information that allows them to make better judgments on value. Thus, restating assets to reflect their current value seems like a good thing to do. I am not sure that accountants are in the best position to make this judgment or that balance sheets should be constantly restated to reflect the accounting estimates of value, and here is why: 

1. Accountants already have plenty to do in terms of estimating earnings, debt outstanding and capital invested. Adding one more item to their to-do list can be a distraction. 
   
2. By their very nature, accounting estimates of value have to be based on clearly defined rules and standards to prevent game playing. That works well for conventional accounting but not for valuation. For every rule in valuation, there are dozens of exceptions and it is impossible to write a FASB rule that captures the exception. 
   
3. The very notion of fair value is a nebulous one, since the fair value of even the simplest assets can vary depending upon what parameters you put on it. For instance, the fair value of a company run by its existing managers can be very different from the fair value of a company run optimally. Similarly, the fair value of a private business for sale in a private transaction can be very different from the fair value of that business to a public buyer. 
   
4. Illiquidity is a wild card in the entire process. Traditional valuation models capture the intrinsic value of an asset, but what someone is willing to pay for that asset will reflect the illiquidity in the market. The problem with pricing illiquidity is that it can not only vary across time but also across investors. A long term investor with a substantial cash cushion, will care less about illiquidity than a short term investors, and all investors care more about illiquidity during crisis.
   
5. Marking to market is applied inconsistently across asset classes. For instance, marking to market seems to followed more religiously when it comes to security holdings than it is with loan portfolios. Thus, financial service firms that have securities on their balance sheets seem to be held to account but banks with bad loans get a pass. 

So, is there an intermediate solution? I think accountants should steer away from estimating the fair value of assets that are long term assets; let's dispense with this move towards balance sheets reflecting the values of brand name, customer lists and other such assets. Fair value accounting is an oxymoron: what you will end up with will be neither fair value nor accounting. With securities that are held for trading/sale, I agree that we should have a different standard but rather than reflect what firms would get for those securities today in the market (which is a liquidation value), I would suggest that firms  either provide estimates of intrinsic value (or the raw data that will allow investors to make that estimate themselves). For mortgage backed securities, as an investor, I would like to see what types of mortgage backed securities are on the books of these companies, what the promised cash flows on the mortgages are and the default risk that they face.