Viewpoint: The European Central Bank’s Latest Measures

Impact of the measures
What will the new policies do? Take, first, the targeted long-term refinancing operations. The program will likely lower borrowing rates for some banks, especially those with weaker balance sheets in the periphery, but I believe the credit that this program can create in the short term is modest. First, because the allotted €400 billion isn’t much, relative to the size of the eurozone economy. The net size of the program will be even smaller if the ECB allows the existing LTRO program to expire at the end of 2016, triggering repayments. Second, TLTROs are designed to be a long-term cure, not shock therapy. They will be introduced over two years, and the ECB loans will have a maturity between two and four years.

Another reason the “teltros” may not succeed is that banks are busy toning up their capital ratios. In the eyes of bank regulators, new loans bring additional risk to the banks’ portfolios. To get in financial shape banks are trimming high-risk assets. Deleveraging has been intense in Spain and Italy, where some banks face pressure to pass the next round of stress tests. Businesses in those areas are thus least likely to find a bank willing to lend. Cheap funding from the ECB does nothing to eliminate credit risk, and so I think TLTROs won’t stimulate credit growth overall.

On the whole, supply isn’t what limits fresh lending in Europe. Most banks can now access funding in the open market. Interbank lending spreads are within the normal range. A few institutions with problems to issue debt, mostly in the periphery, will find the TLTRO terms attractive— assuming they can find borrowers to whom they’re willing to lend. But supply-constrained markets are a minority. More likely is that, overall, the absence of credit growth is due to weak demand.

The closest benchmark to the new lending program is the Bank of England’s (“BoE’s”) Funding for Lending scheme (“FLS”). Launched in August 2012, the FLS allows commercial banks to borrow from the central bank at discounted rates, as low as 0.25%, against collateral. The borrowing limit is given by the amount of net lending banks do to nonfinancial businesses during a reference period. The interest rate on borrowings from the FLS increases if banks reduce their portfolio of loans.

The BoE’s experience hasn’t produced clean evidence of what “targeted funding” can do because we can’t conduct a what-if experiment –i.e. what if the BoE had not started FLS? An evaluation of the program based on observed, raw outcomes may not be valid. That said, available analyses of the still-young Funding-for-Lending scheme suggest there’s been a reduction in banks’ funding costs, mortgage rates, and business loan rates.

The program so far has failed, however, to increase bank lending, especially to small and medium enterprises. The program was ineffective at increasing non-mortgage credit, and only somewhat successful at raising mortgage lending—which is the type of debt the TLTROs exclude.

The three largest banks in Britain (Royal Bank of Scotland, Lloyds, and Santander) have largely ignored the Funding for Lending Scheme and focused on shrinking their balance sheets and meeting higher capital requirements. For banks, like those three, that reduce loan portfolios the FLS charges a penalty that raises the interest rate up to 1.5%, much higher than market rates. Given the goal of sprucing up balance sheets, and that high interest rate, these institutions have little incentive to take up FLS funds.

Consider now the other splashy announcement by the ECB: a negative deposit rate. Charging for keeping deposits, instead of paying interest as is customary, should discourage banks from holding excess reserves. What the measure fails to do, however, is getting the banking system to “lend out” cash balances.

When, say, Bank One uses reserves to buy another asset or to make a loan, it directs the ECB to transfer balances to the counterparty’s bank, Bank Two. Bank One’s deposits at the ECB go down, and Bank Two’s go up by the same amount. Open-market operations by the central bank, as well as currency withdrawals, do change total deposits. But lending and portfolio decisions by the commercial banks don’t.

Lowering the deposit rate, however, does shift down the entire yield curve, through the familiar process. If three-month Treasury bills yield 0.25%, then Bank One would want to get rid of its ECB deposits, now costing ten basis points, and buy some bills. Say Bank One buys bills from Bank Two and transfers ECB deposits to them. But now the central bank charges for those deposits, pushing Bank Two to buy three-month bills too. Banks One and Two bid up the price of T-bills and the yield moves down to -0.1%—plus any liquidity, risk, or transaction premium there might be.

The trading chain doesn’t stop there. Banks trade up the yield curve to take advantage of higher yields at longer maturities, pushing the entire curve down to -0.1%—again, plus the term premium. Arbitrage bids up non-Treasury assets as well, lowering the entire scaffold of interest rates and expected returns.

Throughout this process aggregate deposits don’t change, because the central bank doesn’t engage in asset purchases. The velocity of excess reserves might go up, as banks scramble to avoid the negative yield, but the level doesn’t.

Banks can avoid the negative deposit rate by withdrawing physical currency from the ECB. That’s a profitable strategy, however, only if transporting, storing, and safeguarding currency cost less than ten basis points. In fact you could make money by offering these services to banks at the cost, say, of nine basis points. (This case of arbitrage, though unlikely, illustrates a theoretical point: the lower bound on central bank deposit rates isn’t zero, but some negative number.) Considering the costs involved, however, I don’t think any bank will consider converting their ECB deposits into currency—even king-sized mattresses can’t hide billions’ worth of bank notes.

So far, then, we have determined two qualitative implications of a negative deposit rate. One, interest rates and expected asset returns will go down, possibly touching negative levels at the shortest end of the Treasury curve. And two, there won’t be any change to aggregate reserves. Lower interest rates can, of course, result in an expansion of credit. But that’s a move along the lending curve, thus proportional to the change in interest rates, not a shift of the supply curve of loans.

How big can the effects be? Not much. First, the deposit rate changed by just ten basis points. In addition, there’s no reason why a cut from 0% to -0.1% should matter more than a reduction from 0.25% to 0%, which the ECB did in December 2011 to little effect. Second, total excess reserves are small, around €145 billion. Holding on to those reserves would cost less than €145 million, spread over hundreds of banks. Earnings and capital would scarcely be affected. That banks will raise interest rates on borrowers to make up for the ECB’s “tax on banks” is, then, implausible.

Denmark’s Nationalbank and Sweden’s Riksbank are the only two other central banks that have adopted negative rates before, and neither provides clear evidence of what happens under the zero bound.

You can click here to read the complete report.