Wednesday, March 10, 2010

Changing The Statutory Reserve Ratio: Why It Doesn’t Really Matter

I thought I’d go over this because there seems to be a lot of confusion, particularly from people who should know better. Malaysia is not China (which has raised theirs to crazy-high levels), nor is it Taiwan (currently contemplating a rise in lieu of an interest rate hike).

First, a refresher.

We have to recognise that the modern banking is based on a fractional reserve system, in other words banks are able to lend out money and cash because they know that at any given time, their depositors will only need or demand a small fraction of the deposits the banks are holding on their behalf.

So a reserve ratio of 1% (as it is now in Malaysia) means that banks could in theory lend out 100 times their holdings of cash (read this post for a fuller discussion). Reserve requirements are therefore a prudential measure to ensure that banks have enough cash on hand to meet customer demand – loans are of course illiquid and cannot be used for this purpose, so you have to have some cash or cash equivalents (like NIDs/NCDs).

In a modern financial system, bank reserves are kept at the central bank, where they earn nothing. From a monetary policy perspective, a central bank could therefore theoretically control the amount of credit granted by the financial system through varying the reserve ratio – an increase in the reserve ratio reduces the amount a bank could potentially lend, and vice versa. For instance, going from 1% to 2% effectively reduces the ability of banks to lend from 100x to just 50x their cash holdings. Note that all this happens on the asset side of a bank’s balance sheet.

That’s the economics narrative anyway, the one taught in textbooks.

The problem is, if you have anything like a fair-sized debt securities market (which Malaysia does – the biggest in the region) and an active interbank market, changing the reserve requirement doesn’t really affect lending much. What a change in the reserve requirement will do however is change the composition of banks’ other assets and crimp their interest margins, but not necessarily the growth in their loan book. To see this, let’s have a look at the asset side of Malaysian commercial banks’ aggregate balance sheet (unaudited December 2009 data):


RM billions



Stat Reserve


Deposits placed and Reverse Repos


Amounts due from within Malaysia


Amounts due from outside Malaysia


Negotiable Instruments of Deposits (NIDs)


Malaysian securities


Loans and Advances


Fixed Assets


Other Assets


Total Assets


Note that loans and advances account for just 57% of total assets – you can immediately see the problem relative to the stylised narrative I’ve described above. Since over a third of banks’ aggregate balance sheet consists of liquid or semi-liquid assets, banks won’t have a problem meeting any putative rise in the reserve requirement without compromising their ability to lend.

A hike in the SRR from 1% to 2% will take approximately RM5.7 billion from the banks by my calculations, based on the December numbers. To meet that requirement, banks can draw down their interbank deposits (captured under amounts due in Malaysia) of which there’s about RM30 billion – the rest of it is mainly taken up by BNM repos.

Given some advance warning they can also liquidate some NIDs, or sell down some of their holdings of Malaysian securities, the bulk of which is illiquid private debt securities, but which also include some RM54 billion in MGS for which there is a ready market. But any or all these actions will likely cause interbank rates to start rising relative to the OPR as liquidity gets tighter, which BNM is likely to counteract by pumping in liquidity – which makes the whole exercise moot.

Raising the SRR might raise the cost of funds (from which the Base Lending Rate is calculated), and thus the cost of borrowing – but not by very much, perhaps 1-2 bp per 1% increase from memory. The SRR will only become even marginally effective as it starts getting higher – much higher than even the 4% level that prevailed before the Great Recession started.

The historical high was reached in June 1996, when the SRR was raised to 13.5%, but even then it didn’t put much of a dent on runaway lending as loan growth averaged over 20% in log terms between 1995-97. The same reasons applied back then too, with loans around just 60% of total assets.

I’ve noticed that China hasn’t been very successful in reining in lending using higher reserve requirements. Don’t look for it to be a big factor in Malaysia either.


  1. Nice work, keep it up.

    I was one of those 'people who should know better'. But not because I thought that raising SRR would make an impact (I dont for the same reasons as you).

    Given that the issue here according to Zeti is 'normalization to avoid imbalances'/exit from emergency measures, I thought raising SRR would be the better move (than raising OPR)as a signal to the market that emergency conditions no longer exist rather than raising OPR given that global recovery is still uncertain (to me at least).

    I was wrong.

  2. LOL, I was actually thinking of all those foreign broker guys, who keep lumping all Asian markets up together.

    The signalling reason is a valid one but given that it's a relatively empty gesture coupled with the very broad hints the Governor was dropping, raising the SRR isn't a strong enough start to "normalisation" compared to an OPR hike.

    Having said that, I wouldn't be surprised if there is an SRR hike at the next MPC meeting, in lieu of another OPR hike.

  3. I am no economist, but I admit, I used to do that too. I guess its much easier thinking in blocks rather than each country individually. Now (or rather 2 years ago), I've learnt my lesson!