Removing the securities and loans with the greatest uncertainty from its books would allow the nationalized bank to go forward on a cleansed basis. Once these problem assets were identified and removed, management at the remaining “good bank” could focus forward. Further, potential buyers of all or part of the bank would not need to worry about hidden losses in these assets. At the same time, specialists in distressed assets could be added to those already at the bank and given the chance to work out the bad loans and securities in as effective manner as possible.
The good bank/bad bank structure makes a great deal of sense for these large banks, but it is not a panacea. For one thing, it now appears that perhaps two‐thirds of credit losses will be from categories of loans and securities that one would not normally label “toxic.” For example, it now appears that commercial and industrial loans will have a loss ratio of 4‐7%, probably towards the higher end. This does not sound large, and certainly does not indicate the kind of underwriting, pricing, and transparency problems that existed with the toxic assets. Yet, there are such large volumes of these loans that a rate in that range still creates quite a large absolute loss.
It would be relatively easy to determine which securities represented toxic assets and the same is true for parts of the loan book. Other parts of the loan book may be harder to differentiate. One of the things about a severe recession is that it tends to produce surprises, bringing down some companies that had seemed quite safe and hurting some loan categories more than one would have expected.
Skeptics may wonder why large banks have not already split into good banks and bad banks. Citigroup has in fact already done something similar, but without taking the critical, defining step of setting up a separately capitalized unit that is not supported by the capital of the good bank nor has a mutual parent who is responsible for maintaining sufficient capital at the bad bank. For a bad bank to be effective in freeing the good bank to move forward it has to truly stand on its own.
It may be that the extra benefits of a truly separate bad bank are weaker than appears to be the case, perhaps because the internal management of these assets already lies with specialists who do not operate much differently than they would in a truly separate bad bank. More likely it is because a separate bad bank requires separate capital, which effectively crystallizes the value of losses for everyone to see. That is, the separate bad bank will have to have enough capital to reassure regulators and all other relevant constituents that it can indeed absorb the losses that may still exist in the toxic assets. However, the good bank’s stakeholders are unlikely to give it any credit for possible overcapitalization at the bad bank, even if it legally has some residual upside. Therefore, any capital moved into the bad bank represents a capital loss to the good bank, unless it is obtained from new outside investors in the bad bank, who are likely to strike a very hard bargain for supplying that capital.
In sum, from the good bank’s point of view, moving the bad assets to a separate unit would look very much like keeping those assets and writing them down to their lowest reasonable valuation. The capital hit would be very large and could very well be much larger than the economics would warrant, if there is indeed a significant element of “fire sale” pricing in today’s highly illiquid market for these assets.