Negative interest rates are just a tax on reserves that lowers net interest margins

The primacy of monetary policy continues unabated as central banks go further and further down the rat hole of increasingly desperate measures to boost demand. First, it was quantitative easing. Now, the latest scheme is negative interest rates. They tell us that monetary policy is not exhausted and that still more policy initiatives lie ahead, particularly helicopter money. However, we should be sceptical that any of these policies will gain meaningful traction before another economic downturn. Brief comments below

I have written a decent amount in the past about so-called quantitative easing. And in recent weeks, I have also written – with a large degree of scepticism – about helicopter money. And while I think helicopter money has some merit, if administered appropriately, negative interest rates offer no benefits at all from where I sit. The following chart from Bloomberg sums up the problem with so-called negative interest rates.

Net interest margins

The lower the headline interest rate is, the lower bank net interest margins are. And when the central bank actually taxes reserves, the net interest margins are even lower. So what central banks are calling stimulus is really just a tax that erodes bank capital. Ceteris paribus, this results in less lending, not more; it is anti-stimulative rather than stimulative.

One thing to think about as we review our paradigm on how the economy and the credit markets interact is whether lower rates are stimulative. I have long believed the common wisdom that lower interest rates are always stimulative is too simplistic. And I think what we are now witnessing is an understanding that – in the current situation – lower interest rates are deflationary, anti-stimulative. So I wil use the rest of this post to explain why.

Back in 2010, I wrote a post about how quantitative easing and permanent zero are toxic to bank net interest margins. What I wrote is that in recession “banks will be set up for huge loan losses despite recent under-provisioning. Meanwhile they will have no way to make it back on net interest as long rates come down in a recession while short rates remain at zero percent, killing net interest margins.”

See, this is exactly where the conventional wisdom on interest rates is wrong. First, interest paid is also interest income. So, cutting rates cuts the payments to some while also cutting the interest owed by others. The primary question regarding whether they stimulate the economy goes to whether relief for the debtors or the lenders is best-positioned to sustain economic growth. In a recession, there are huge swathes of the economy that are experiencing debt distress as income erodes. Interest rate cuts precisely at this time allow enough of these debtors to make payments on time and in full to prevent the knock-on effects that come from defaults that reduce lender interest income.

The real problem with deflation is its genesis in demand shortfalls from the debt deflation caused by lost income that lead to borrower defaults and a chain of further knock-on defaults due to the lost interest and principal repayment income of creditors. We saw this last cycle when the banks were insolvent on a mark-to-market basis due to the losses they took on mortgages. it was only the government capital injections, the suspension of mark-to-market accounting, and the Fed’s QE1 intervention in indiscriminately buying up MBS paper that prevented a full-scale Great Depression scenario.

Second, once the economy recovers and the low interest rates remain in place, they cause net interest margins to suffer because interest rate expectations cause the yield curve to flatten. The way the Fed exerts a dominant influence across the yield curve, not just on the short end is by signalling where rates will be in the future or – more precisely – allowing the market to gauge where it believes rates will be. As I put it in 2013, “markets know, therefore, that the Fed, as a monopolist, will always be able to hit its Fed funds target now and in the future. Therefore, future overnight rates reflect only future Fed Funds target rates as set by the Federal Reserve. This means that future expected overnight rates reflect only market-determined median expectations of future Fed Funds target rates as set by the Federal Reserve (plus a risk premium). Long-term interest rates are a series of future short-term rates”.

If the Fed signals that rates will be lower for longer, unless the market sees something that makes this signal not credible, rates will decline. And in places like Japan, Switzerland, Denmark, Sweden, and the Eurozone, markets are signalling both an understanding that rates will be negative or low for the foreseeable future as well as a preference for duration via term premia. That means yield curves in those currency areas are as flat as a pancake and net interest margins reflected in the first chart are correspondingly low.

Finally, interest income is an important source of income for savers in the private sector. Take this income away and you could get perverse behavior due to a loss of income, meaning people save more rather than less due to the fall in rates.

The key here is to realize that a policy of cutting rates during an economic slowdown in order to diminish debt distress, defaults and the resulting knock-on income losses is not the same as keeping rates low permanently. When the economy is recovering and debt distress is diminshed, the negative impact of falling net interest income for banks and falling interest income for households and businesses works as a tax, reducing income and, therefore, slowing outlays and the economy.

Central banks, with their negative rate policy in the middle of a recovery, are essentially taxing banks. And this tax can only reduce lending unless banks take on risk and duration. And there are signs they are doing so. Think about the 50-year Spanish government bond and the 100-year Irish government bond that were issued this past year. These are countries that just a couple of years ago were in Troika-administered debtor programs. Yet, Ireland got 100-year paper out the door at 2.35% yield. And last year, the basket case that is Petrobras also issued a 100-year bond. Serial defaulter Argentina came back to market this Spring with $15 billion dollars in issuance that was more than 4 times oversubscribed, at yields lower than marketed because of the high level of investor interest.

In short, low and negative nominal rates cause investors to reach for yield, to take on risk and duration in a way that may cause their returns to suffer when the economy turns down. At the same time, these policies work perversely because they reduce interest income at a time when a lack of debt distress reduces the need for a positive impact on debtors.

I expect central banks to continue with this experiment though. And the negative rates will fail to boost the economy, except through portfolio preference shifts that misallocate capital toward duration and risk. When the economy turns down, rates will still be near zero and the yield curve will flatten, not steepen. Banks will have zero net interest buffer because of the flat yield curve. And the full weight of increased loan losses will come to bear on their balance sheets. The result will be a severe shortfall in credit growth, a deepening of the downturn and a bear market in risk assets.


Edward Harrison is the founder of Credit Writedowns and a former career diplomat, investment banker and technology executive with over twenty years of business experience. He is also a regular economic and financial commentator on BBC World News, CNBC Television, Business News Network, CBC, Fox Television and RT Television. He speaks six languages and reads another five, skills he uses to provide a more global perspective. Edward holds an MBA in Finance from Columbia University and a BA in Economics from Dartmouth College. Edward also writes a premium financial newsletter. Sign up here for a free trial.