David R Kotok
August 16, 2016



The bottom line is that we think the financial/banking system is tightening and we are now defensive. We have raised cash in our US ETF accounts, and we have taken bond profits and shortened duration in total-return bond portfolios.

We are applying very high standards to credit. No junk bonds. No opaque fund-of-funds-type hedge funds. Nothing with a lockout period. If you cannot turn it into cash and you cannot get your money whenever you want it, we won’t touch it.

Here’s why. This is going to take readers a few minutes, and some of it may be a little technical.

The basic assumption is that there are two types of monetary tightening. The first is visible: rising real interest rates. The classic test is that you see the nominal interest rate rising faster than the inflation rate or the expected inflation rate. This kind of tightening is usually administered by a central bank, as it increases the policy-setting interest rate.

The second form of monetary tightening can happen without changes in the policy-setting interest rate. It occurs because the rules change. It is harder to see, but it is just as powerful as the first form – it may even be more powerful. We believe the second form of tightening is under way right now.

Rules have changed for the big banks worldwide. The former paradigm of huge, unrestricted excess bank reserves deposited with the central banks has given way to the absorption of those reserves by Basel III rules regarding High-Quality Liquid Asset (HQLA) tests and Liquidity Coverage Ratio (LCR) tests. We have written about that in the past. (See For superb research on this subject, see the work of Zoltan Pozsar and Sarah Smith at Credit Suisse.

Now we have a set of rules changing the nature of money market funds. There will be two types of funds. One will not “break the buck,” and it will hold short-term paper similar to the HQLA 100% test. The other type will likely pay a higher interest rate and will hold paper not meeting 100% under HQLA rules. The second one includes commercial paper (CP). The rules on money market funds take effect in mid-October. Market agents have been moving toward them for the last 6–10 weeks.

In anticipation, some money market funds are closing or will close. Others have morphed their business books into bond funds. The financial press has been filled with stories of this change. What the press hasn’t discussed is how this change in flow has raised bond prices through the fund flows.

Now we see the impact of the money market rule change. LIBOR is rising. It reflects the interest rates that are impacted by the non-HQLA subset. Ties to LIBOR are also rising. We can see this in LOIS (the LIBOR-Overnight Index Swap spread). We can see it in the TED spread (the difference between Eurodollar futures and US Treasury futures). We can see it in the cross-currency interest rate swaps. (A cross-currency swap involves converting a floating rate to a fixed rate with a swap and also doing one side in one currency and the other side in another currency.)

Let’s stop the acronyms and leave the details to professionals. But before we do we want to recommend Donald J. Smith’s research paper 2012-11 at the Boston University School of Management, dated June 1, 2012 and entitled “Valuing Interest Rate Swaps Using OIS Discounting” ( We also want to recommend the October 3, 2012, paper by Jawwad Farid entitled “Using OIS Swap Valuation – Overnight Indexed Swap Rates versus LIBOR” (

Okay. So we will try to put all this in plain language.

The spread between the highest-quality shorter-term securities like the interest rate on Treasury bills and other non-HQLA instruments is widening. Market agents are saying that phenomenon is temporary due to the restructuring of the money market funds. They may be right. But we think they may also be wrong. We are not sure. And we cannot find anyone else who is. There are many opinions and assertions, but there are no facts.

What we do know is that the rates in LIBOR and related metrics are rising faster than the central bank’s policy rate. Example: the Fed has raised its policy rate one time and by 25 basis points during the last year. Meanwhile, 3-month LIBOR was about 30 a year ago and is now over 80 and rising. LOIS was 20 basis points two months ago; it is now over 40. “Japanese banks face dollar funding pressure,” Joe Abate of Barclays observed on August 11. That pressure is about $100 billion in size. Australian, Canadian, Swiss, and US banks fund themselves with CP in the hundreds of billions.

So all these rates are rising, which means that the cost of funding to lending institutions is rising, as is the pass-through of that cost to the borrowers.

Here is where it gets harder.

No one knows how much debt and derivative exposure is tied to LIBOR or the related metrics. We have seen estimates as high as $28 trillion worldwide. Others estimate that about 15%–20% of US household debt is directly or indirectly tied to LIBOR. There are other significant numbers for commercial debt. And the derivative total size is huge but really unknown.

Let’s use $28 trillion.

Each single-basis-point rise translates to $2.8 billion in additional annual interest cost imposed on the collective worldwide dollar debt by this systemic change. So, annualized, $56 billion a year is the rate of change from only two months ago if we use LOIS as the metric. If we use 3-month LIBOR as a metric and the last year for analysis, we see a change of over 50 basis points or about $10 billion a month. Furthermore, we estimate that these numbers will go higher. Some think the trend will reach the level where LIBOR exceeds 100 and where the foreign banks in NY will need to use their central bank swap line option. That kicks in at about 120.

The complacent side of the market seems to be dominant today. Thus stocks hit new all-time highs, and policy debates focus on when the Fed may move the rate up (best guess is a 50-50 chance in December). But the spreads are telling us a different story. They are sounding alarm bells, and those bells are getting louder.

At Cumberland, we respect spreads. We watch them daily. We see the gap widening between riskless and risk. And we see new rules absorbing the excess reserves that were created by the central banks in response to the last crisis. And we are worried about the next crisis.

So we took some action in the interest of preservation of capital and safety. That is why we have a cash reserve in our ETF accounts, and that is why our bond portfolios are increasingly defensive.

We have a postscript. Readers may not know that we often vet writings internally. In this case there were some critical and helpful observations.

My colleague Bill Witherell noted (citing Ned Davis Research) that we are not seeing a similar LIBOR response in other countries and in other currencies. That is evidence of the money market fund shift being the cause, since it is US-centric. One might also point to the negative-rate regimes altering behaviors in ways that would keep this LIBOR tightening phenomenon from appearing in those currencies and countries.

My colleague Bob Eisenbeis had added observations. He notes that central banks can raise rates as a policy tightening. The second, rules-change, tightening we discussed above is really a change in market supply and demand conditions that increases the demand for short-term liabilities relative to their supply. In this case, the increase in demand is due to regulatory changes motivated by financial stability concerns (which may or may not reflect a conflict between those concerns and monetary policy objectives). Bob argues that there are two forces at work here that need to be spelled out in a bit more detail. The demand for reserves and HQLA should bid up prices and put downward pressure on rates. This reflects an increase in demand, relative to supply of those assets. The increase in rates and, we note, on spreads seems to be risk-related since institutions not holding the HQLA must now pay a higher rate because of the risk that they could break the buck. Bob asks whether such activity is quantitatively significant enough to cause such a widening of the spread. Clearly the affected money market funds have to be small relative to the $28 trillion of credit tied to LIBOR. Bob suggests that if our position is correct, then macro prudential policy is working at cross purposes to monetary policy.

What's been said:

Discussions found on the web:

Read this next.

Posted Under