I used to work as a consultant on banking compliance projects. I chose this area because it promised to be complex and therefore interesting, and because projects of this sort were unlikely to be canceled: the banks were required to do this work. Unfortunately, they were not required to do it well: in pre-crisis days, so long as they were doing it with some show of sincerity (i.e., so long as they were throwing a lot of money at the problem) actual progress toward the goal of measuring accurately the risks they faced, actual progress I say, was not necessary. The banks were fantastically profitable and politically powerful, accounting as they did for an undue proportion of employment and prosperity generally. The ability of those charged with "regulating" the banks to punish or even persuade was very limited (shown below, the view from my office window with De Nederlandsche Bank--the Dutch regulator--off in the distance, near the tip of the rainbow, as though it existed in some fantasy world). However, by international agreement, it had been decided that most of the world's more important banks would start measuring their risks in roughly similar ways, and that eventually they would report publicly on those measurements. Hence these projects.
"Measuring the risks," that was the key to the whole thing. Banks, like casinos, are machines for generating profits. They do this by borrowing money at a low rate of interest (for example, from the government, or from you in the form of a savings account) and by lending it out again at a higher rate (casinos, of course, just take your money, while making a few falsely suggestive pay-outs). This simple model works so long as two important things hold true: first, the general structure of interest rates must hold relatively steady such that what is "low" stays "low" and what is "high" stays "high," and second, borrowers must pay back what they have borrowed, or most of it anyway. (There are lots of less important things that need to hold true, too--for example, the bank's employees must not steal all the money--but these two are the real keys.) Everything the bank does involves risk--"interest rate risk" (low is low, high is high), "operational risk" (the bit about employees stealing), and many other kinds--but how a bank manages credit risk (i.e., the risk associated with repayment) is probably the one thing that most clearly differentiates the good banks from the bad.
It was in credit risk that I did most of my work. It was the most interesting area, to my mind, and it was the most active one, too, because the regulations that require banks to sit on some of their money weight the threat of defaulting counterparties more heavily than any other risk the bank faces. One effect of this is that some banks have come to view credit risk reduction primarily as a means of reducing the amount of money they are forced to keep on the sidelines. Credit risk can be controlled by being more selective in your choice of counterparty, by investing in research, by making conservative decisions, by sticking to markets and products with which you have long experience, by being what most people still think of as a "good banker." But the credit risk figures can be reduced much more easily than that. They can, in a word, be fiddled.
Fiddling your figures--that is to say editing the numbers you report by inserting adjustments of one sort or another into the stream of calculation--is a very much more efficient means to this end. Not only is it pretty easy to do, but, most important of all, it doesn't require you to change the way you do business. (Hey hey, don't you get all judgmental here: if the way you did business was earning you a billion bucks a quarter, you wouldn't want want to change it either!) And that's what some banks did: they fiddled their figures, more or less with the collusion of the regulators (OK, now you should get judgmental), while claiming to be improving their calculation systems such that those systems would soon report the "truth," which truth, they claimed, would turn out to be the same as that shown by the "corrected" figures.
Now I'm not saying the banks I worked for did this. No, no, I'm not saying that. And even if I were it would be kind of a nonsensical thing to say, because the banks I worked for were enormous multi-faceted organizations that were often internally conflicted and where they weren't were as like as not in a situation where the one hand knew not what the other was doing, so to tell it as if "the" bank behaved in a consciously fraudulent way is just not quite right.
All I'm prepared to say in this forum is that because of the way things were set up--because regulators were weak and banks strong, because everyone was high on the pipe dream of infinite profits, because most of us despite it all still did and do respect and trust financial "institutions"--there was so much of this going on in the world of compliance that if you actually cared about getting these calculation systems to work correctly you would probably feel rather out of place a lot of the time. You would repeatedly find yourself much more concerned about the dysfunctionality of the process of development than were those ultimately responsible for that development. You would find it increasingly difficult to support the goal of being declared on a sound footing when the steps you witnessed day to day were downright drunken. You would, much to your surprise and as testament to your immaturity, discover that you were an idealist, and that this was no business for idealists.
