The one thing we can’t accuse central banks of these days is lack of creativity. The latest gem came from the Central Bank of Turkey (CBT) last week, when, on one hand, it cut its policy rate by 50bps to 6.50%, while at the same time increased the reserve requirement ratios (RRR) for short-term bank funding (deposits and repo) to help lengthen the maturity structure of banks’ liabilities.
I don’t want to dwell exclusively on the Turkish example, which, in my view, is fraught with confusion about what exactly the authorities are trying to achieve. What I do want to do is examine under what conditions, if any, a hike in reserve requirements can be effective in tightening monetary conditions. This is particularly relevant at a time when many emerging markets think they can “get away” with avoiding raising interest rates by employing alternative tools, to avoid attracting further capital flows from abroad. Changes in the RRR are meant to influence monetary conditions through the so-called bank-lending channel of monetary transmission. Accordingly, bank funding relies in large part on demand deposits, which are subject to reserve requirements. Raising the RRR increases banks’ demand for reserve deposits at the central bank (CB). If the CB does not accommodate that demand, short-term interest rates will rise to bring demand for reserves in line with supply. Monetary conditions then tighten because banks will pass on their higher funding costs to corporates and households, in the form of higher lending rates, in order to safeguard their profitability (interest rate margins).
Against this backdrop, for a RRR hike to be effective in tightening monetary conditions, the following have to hold:
Banks should be limited in their ability to switch to other sources of short-term funding that are not subject to reserve requirements. In the case of Turkey for example, the RRRs (which are different across different liability maturities) have been applied to both deposits and repo funding from abroad and from domestic customers. However, they do not apply to repo transactions with the CBT and among domestic banks. For this reason, it’s unclear how the hike in RRRs can materially increase banks’ cost of short-term funding, if the CBT is effectively committed to supplying enough reserves to keep its target rate at (the now lower) 6.50%.
To make things more tangible, suppose a bank has 100bn liras worth of short-term liabilities and the RRR is raised from 6% to 8%. Suddenly, demand for reserves at the CBT rises by 2bn liras. Banks would then have to sell 2bn of their other assets, to meet the requirement, putting upward pressure on interest rates and, in the process, also shrinking the quantity of credit in the economy. However, by committing to maintain the target interest rate at 6.50%, the CBT effectively commits to creating enough reserves to meet any additional demand at that rate.
How will it do that? It will go to the open market, purchase TRY2bn worth of government securities and fund the purchase with the creation of bank reserves. Alternatively, it can go to the FX market and purchase TRY2bn worth of, say, US dollars, again funded with bank reserves. In that way, banks’ reserves go up without them having to shrink their loan book. On the contrary: new liquidity has come into the system (as the cash received by the sellers of the bonds or the foreign exchange has been deposited at the banks), which can be employed for further loan creation.
Note that the measure might still be desirable from a prudential perspective—e.g. to the extent that higher RRRs on foreign repo funding might encourage a shift towards lira-denominated funds, thus avoiding undesirable currency mismatches on banks’ balance sheets. But what I’m arguing here is that the RRR hike is unlikely to restrain credit growth because (a) it would not have a material impact on the price of credit; and (b) it might even contribute to increasing the quantity of credit, if the CB commits to accommodating whatever additional demand for reserves at the going 6.50% rate (via securities purchases or unsterilized FX accumulation).
Looking beyond Turkey, my bigger point is that measures that do not materially raise the cost of capital (and, indeed, that are intended to lower it, to fend off foreign capital inflows) are unlikely to tighten monetary conditions. Not only because their impact through the bank-lending channel will be diluted; but also because they ignore other important channels of monetary transmission—notably the wealth and financial accelerator channels.
For example, if short-term interest rates remain low, cash will seek riskier assets, boosting their prices and leading to positive wealth effects. This in turn triggers pro-cyclical investment and consumption and, potentially, overheating (the wealth channel). Admittedly, when capital markets are global, low foreign interest rates can also play a big role in boosting asset prices, so an increase in the local short-term rates might not be enough to generate the desired tightening.
Similarly, rising asset prices increases the value of collateral and thus, the perceived creditworthiness of borrowers, encouraging more lending (the financial accelerator channel). Higher RRRs on domestic banks, IF binding, could restrain the extent of such an increase, but even then, in an open economy, at least part of the capital “shortfall” is likely to be covered by foreigners.
Finally, “here and there” policy measures may be ineffective or counterproductive, by creating confusion about what a central bank is trying to achieve and undermining its credibility. Is it lower credit growth to curb domestic demand and a widening of the current account deficit? Is it the promotion of financial stability through the change in the maturity structure and currency composition of banks’ liabilities? Is it the discouragement of capital inflows and the prevention of “excessive” exchange-rate appreciation? What about that price stability?
Clearly, there are no easy answers to the policy dilemmas of our times. My suspicion is that we should be bracing for blunter measures in the year ahead.
Originally published at Models & Agents and reproduced here with permission.