The FSA missed the liquidity problems at Northern Rock and Bradford & Bingley. certainly at the latter that was the entirety of their problems although the former had problems with the quality of its assets. But we can be sure that liquidity risk wasn't the FSA's strong point, and by their own eventual admission, they goofed. They thought that the Bank of England would step in if required so they left all the monitoring to them. The BoE quite rightly said that the FSA wanted all bank regulation and supervision to fall under its empire, so it was the FSA's responsibility.
So when the FSA found out they had to do something about "liquidity risk" they ran off to their copy of the Beginner's Book of Financial terminology and looked up "liquid", and "liquidity", and they pretty soon came across the term "mandatory liquid assets", which in the UK were non-interest bearing deposits with the Bank of England that banks were required to hold on their balance sheets. The term was a bit of a misnomer because assets would have been liquid but for the fact that they were mandatory, or to put it more clearly, the banks could have raised ready money by withdrawing the deposit, but it wasn't allowed to.
In fact the mandatory liquid assets had little to do with liquidity and much more to do with monetary policy and constraining bank lending. By increasing the proportion of assets that had to be held on deposit banks could be dissuaded from high volume, low margin lending and over inflating the money supply. That is nothing to do with liquidity risk, which is all about a mismatch between the duration of borrowing and lending in a bank which can result in a cash squeeze if the market declines to roll over short term funding while assets are outstanding.
So what is going on today? The FSA has said that bank's will have two options: on the one hand they can run asset and liability books where the durations of both side of the balance sheet are broadly matched. This reduces liquidity risk because as assets are repaid, the corresponding liabilities can be repaid and there is no risk of a funding squeeze.
That is a conservative form of banking, but bank treasuries have always made a profit by borrowing shorter term than the bank lends and making money on the average lower interest rates for shorter term borrowings - the old borrow at 3, lend at 4, on the golf course by 5 mentality. But that does come with the risk that when the borrowing comes to be refinanced half way through the asset's term the bank cannot source funds at an affordable price - a liquidity crunch. In practice banks can live with this risk so long as they don't get too aggressive and make all their profits from borrowing short and lending long, which is what killed B&B and Northern Rock.
On the other hand the FSA says if a bank runs an unbalanced book they have to hold 15% up to of their assets in government debt. This doesn't cost the bank anything in terms of capital because the gilts are zero weighted for capital adequacy purposes, so the bank could simply go out and borrow several billion on the money markets to buy the gilts, but it does cost the bank to do so because of the spread between the yield on gilts and the banks cost of borrowing. So the bank will try to mitigate that by borrowing as short term as possible compared to the long term gilts.
But, you might think, the gilts are liquid assets, so the bank can sell them to raise if they are short of funding. Think again, the gilts may be liquid, although you could now expect the price of gilts to fall through the floor if there is a bank funding crisis when all the banks rush to sell. It ois worse than that. If the bank is required to hold 15% of its assets in gilts, then it can't sell them to raise cash, so while the gilts are liquid to the market, they aren't liquid to the bank.
Now the theory is that the bank has a buffer of liquid assets equal to 15% of its portfolio, which in extreme circumstances it could liquidate, albeit that at that time the regulators would start to step in, but my guess is that by stipulating this liquid asset buffer and its associated costs, the banks' attempts to mitigate the negative spread between bank funding costs and gilt yields will mean that they will run a much higher degree of liquidity risk than before and probably increase the overall liquidity risk.
Right problem (finally), wrong solution. Think again, FSA. No better still, step aside and let the Bank of England take over. They know what they are doing.