Friday, 6 June 2008

Show me the money

When I was young, I thought that banks were full of cash. I imagined that my local high street bank had a vault, buried deep underground, holding uncountable millions of bank notes. Then I grew up and learnt about fractional reserve bank.

Far from holding huge quantities of banknotes, I learnt that banks hold a small fraction of their assets as cash. Through a trick of double entry bookkeeping, banks create electronic money, using what little cash they have to manage the daily customer demand for notes.

Of course, fractional reserve banking has been around for centuries. Nevertheless, today things are different. It is hard to imagine a time when banks held such a small proportion of their assets as cash.

For several decades, the bank of England has measured commercial bank liquidity. Three of those measures are presented in the chart above; the broad ratio, the reserve ratio and the narrow rate. The precise definitions of these measures are unimportant. Basically, they all measure the proportion of cash or near-cash financial instruments banks hold relative to their total assets.

Since the mid-1960s, banks have dramatically reduced their holdings of liquid assets. Most of the decline happened in the early 1970s, when financial markets were liberalized.

However, the long-term chart hides a more recent decline. Here is the same chart showing the same ratios over the last ten years. Again, the story is dramatic. At the risk of exaggeration, banks hold virtually no cash or liquid assets.

This problem has not gone unnoticed down at the Bank of England, though there is little evidence that the FSA realize it. Here is what the BoE recently said about commercial bank liquidity:

"Some measures of sterling liquid reserves have fallen since 1967. Banks have diversified their liquid asset holdings beyond those shown — for example, to include other currencies — and increased their use of repo markets, but the overall historical pattern is one of a marked decline in cushions of high-quality liquid assets."

Is this a problem? Again, here is how the BoE frame the issue:

"While firms themselves have strong incentives to be resilient to liquidity shocks, they may make less provision for liquidity risk than is desirable for the system as a whole."

Decoding the sentence I think is saying that there might be a problem here.

Recently, the BoE, along with other central banks and regulators have been looking into the issue. In the wake of Northern Rock and Bear Stearns, liquidity ratios are likely to come back into fashion. It is time for banks to again start holding cash down in the cellar.


Anonymous said...


Do you think electronic payments (eg. Maestro, Visa, SWIFT, wire transfer, Nostro etc) may be a big part of this? If you make lots of your transactions electronically you don't need cash per se, you just use your magical double entry.

Also, efficient electronic clearing at the BoE between banks means a Bank only needs monitor the net position of the entire Bank, not the position of each branch

Just a guess. I'm not sure if that near-cash number includes funds available for the above.


lottery said...

It could challenge the ideas of the people who visit your blog.

aSteve said...

I'm unsure about your conclusion, Alice... I think there are lots of potential alternatives to holding "cash".

Risk management in general needs to be considered very carefully - but, for example, secured lines of credit might be an alternative.

Something I noticed in unrelated news recently was a plan by banks to effect instant electronic payments... cutting out the current 3 or 4 day delay when moving cash between two electronic accounts. Even seemingly trivial schemes like this improve liquidity.

P.S. While you say that the definitions of broad, narrow and reserve ratios are dull, any chance of a link?

Anonymous said...

I think it depends on the type of assets the retail bank holds. If they do a Bradford and Bingley and decide to lend large sums of money to speculators for over priced assets then there is clearly a problem. But all the time the bank has cash sitting around their vault doing nothing then effectively they're losing money as they're their creditors will still expect a return.

Asteve pointed out to me a while a go about recent changes to the BofE procedure which mean that they now pay interest on deposits made by retail banks. This should mitigate some of the problems associated with leaving large amounts of cash idle.


aSteve said...

Dave's on about the 2006 "Reform of Sterling Money Markets." It is only my interpretation that, in future, banks will be able to hold cash without lending it to risky counter parties without loss of interest. There are complications - for example, the BoE expect banks to agree their reserve balances in advance... and there is a definite belief that the reserves deposited at the BoE will be smaller than the sums borrowed. The BoE argue that the changes make its operations more straightforward by allowing weekly rather than daily reconciliation. While starting paying interest at a central bank might, at first, seem to be inflationary, I'm sure that at present it has little effect, and that if, in future, it turns out to lead to larger central bank reserves, it will prove deflationary as banks lend less and save more... hence acting as a brake on M4 expansion.

alice cook said...


the data comes from the BoE. Again, it is the Financial stability report. BTW, liquidity is not just cash. Financial innovation will reduce the need to hold notes and coins, but banks are economising on liquidity, which includes highly liquidity assets such as t-bills. My view is that banks have gone too far. In this respect, I am only mouthing a view held by many central banks, who are looking to tighten up liquidity management.


alice cook said...

asteve, on your second comment, my simplistic view of the 2006 reform was that the boe moved to a superior form of short term liquidity management, which the ECB were using. In effect, the ECB's method ensured more stable overnight interest rates, whereas the older BoE operating system was characterized by higher short term interest rate volatility.


Josh said...

So the banks are holding no cash - great!

aSteve said...

Alice, I agree 110% that banks have failed catastrophically in their liquidity management. My interpretation is that, in the last few years, almost all risk has been ignored. A lack of regulation of the market effectively forced every bank to take ever riskier positions or accept under performance within the sector. The FSA should have been rigorous in their regulatory role, but - it transpires - were worse than useless. I think this has substantially undermined the ability of the BoE to manage the economy as we'd hope.

On the figures, I guess I have to go back and do more reading of the Financial Stability report... it is an interesting document - though it does impose a fairly steep learning curve, I've found.

Your view of the 2006 reform is exactly that published by Paul Tucker - who, as far as I can see, was the man most involved with the changes. The ideas about how reserve accounts paying interest might effect our currency is purely down to my own gedankenexperiment. At first it seemed highly inflationary... but, after careful thought, I think it is deflationary. Paul Tucker's position is that it will be neutral - though he only explains why the effects should be tiny... I've no evidence at all that they have not been, It was something I found interesting several months ago... primarily because it lead to discontinuities in financial statistics... which I thought suspicious. ;)

My own view is that, in the short term, the change will encourage repos that might not have been attempted previously... because, even if the money was not then lent commercially, then it can be kept as a reserve balance to cancel the interest payable on the repo. I think that this (in the context of widening the variety of acceptable repo assets and the SLS) might skew M0 figures slightly - but that there would be no effect on the economy. In future, it might encourage demand for repos - since, with interest paid on reserve balances, it would be impossible to repo "too much" - assuming that reserve balances could be appropriately negotiated.

Cynically, I'd point out that since 2006 we've not had especially stable short-term interest rates - just look at LIBOR for proof of that. I think that the policy change is subtle and so will be its effects.


Never mind the banks reloading their cellars, Karl Denninger wrote recently that WE should be holding cash, in a secure home safe.