By Rob Kirby
Kirby Analytics

The best info on derivatives activity is published quarterly in arrears by the Office of the Comptroller of the Currency [OCC]. The most recent quarterly report was for Q3/20, published in the dying days of December 2020. Thus, our "look back" on derivatives activity for 2020 only provides a "snapshot" of the first nine months of the year. What we now know is that aggregate notional derivatives holdings by the top 25 U.S. Bank Holding Companies has fluctuated from 228.2 Trillion (T) in Q4/19, to 267.6(T) in Q1/20, to 242.8(T) in Q2/20, to 239.5(T) at the end of Q3/20. 

So, it would appear that was a 40 Trillion notional "bloat" in Q1/20 which has somewhat dissipated during Q2/20 and Q3/20. If anyone is wondering how much "joy" 40 Trillion worth of notional derivatives buys these days, take a peek at what happened to the yield on 10-year U.S. government bonds in Q1/2020 below:

Yield on 10-year U.S. government bond – 2020

Interest rates didn’t "fall"; they were "THROWN OFF A BUILDING".

It should be noted that the overwhelming majority of all outstanding derivatives are interest rate products [predominantly interest rate swaps and FRAs (forward rate agreements).

A Few Words About How/Why the Derivatives Heap Got So Big

FRAs are short-dated interest rate derivatives (or bets) that generally trade against (for hedging purposes) 3-month Eurodollar futures because they settle against Libor. But if the Treasury wanted to compel an entity to "engage" in trade—they simply sell them at yields lower than T-bills making bank counterparties willing buyers of ANY AMOUNT of T-bills at any rate—period (because Libor ALWAYS settles at a higher rate than risk free T-bills). FRAs are generally effective and liquid out to 18 months’ duration.

What most laypeople do not fully understand is that interest rate swaps of duration 3–10 years have cash bond trades embedded in them, at least in the case of banks. This is because banks are spread players—managing the spread between the yield on U.S. Treasuries and their own cost of funds—and they are not allowed to take untold Trillions of naked interest rate risk. The U.S. Treasury (specifically, the Exchange Stabilization Fund or ESF) does take naked interest rate risk and serves as the backstop counterparty—receiving fixed—for the bulk of these trades. This hugely gooses transaction demand by making the banks (Commercial or Bank Holding Company) captive buyers of cash U.S. government bonds in virtually any amount the U.S. Treasury wishes.

Using FRAs (short term) and interest rate swaps (3-10 years) leaves a scheme to control the entire interest rate curve with a "hole to fill or a conundrum"—right at the 2-year duration point, as in, how to compel market participants to buy large amounts of 2-year U.S. government debt. Answer to that: Operation Twist, where the Fed effectively exchanged (bought) long-term bonds for short-term bonds—dollar-for-dollar, making 2-year bonds cheap.

If anyone still doesn’t "get it"—that derivatives are primarily used to CONTROL and create predetermined outcomes—please take a look at the following graph (Graph 1: Derivative Notional Amounts by Type) found on page 18 of the PDF of the most recent quarterly derivatives report:

Graph 1 from Q3/2020 Quarterly Derivatives Report

The green line that hugs the horizontal axis at the bottom of the graph represents "end user demand" of all aggregate derivatives. This is, in itself, a STUNNING admission that there "is" no observable demand in the free world. Regulators are loathe to admit that the real end user is the U.S. Treasury’s ESF, and the true intent of this "orgy of interest rate derivatives" is DE FACTO one of the prime conduits which transmits imperialist U.S. monetary policy. Of course, no one in the mainstream financial press dares speak of activities of the ESF, because EVERYTHING the ESF does is, by definition, a matter of national security, and any such utterances make one as popular as Julian Assange or Edward Snowden.

The elevated/artificial transaction demand created by interest rate derivatives gives the U.S. debt complex the ILLUSION of being "deep and liquid," while giving monetary authorities complete and arbitrary control of the interest rate curve. This simultaneously creates the ILLUSION that the U.S. dollar is strong.

In case anyone does not understand the relevance of goosing transaction demand, I would like to offer my explanation as to what has been occurring recently in the crypto markets. In September 2020, the price of gold was 1900.00 per ounce and the price of Bitcoin was 9,000.00. Since then, the price of gold has gone nowhere, while Bitcoin traded above 33,000 earlier today. So, what can possibly explain the complete decoupling? Interestingly, precious metal, which is rare in physical form, can be sold in infinite amounts in derivatives (futures) form on exchanges like COMEX and LBMA (London Bullion Market Association), while virtual assets like Bitcoin cannot be sold more than once due to blockchain protocol. The fundamental difference here being that precious metals futures were "designed" to suppress the prices of the underlying, while Bitcoin was "designed" to emulate honest money.

Remember, China and the U.S. are in a trade war. The U.S. has sanctioned Iran and Venezuela and effectively cut them off from transacting U.S. dollars via the SWIFT system—so, U.S. dollars are of little use to them. So, China, Iran, and Venezuela formed a "club." China buys oil from Iran and Venezuela, and my belief is that they are now purchasing crude in cryptocurrencies from both countries. Iran and Venezuela then recycle the crypto back to China in trade for goods and services they need. This trade completely excludes the U.S. dollar (which is making multi-year lows currently) and has lit a fire under the crypto complex—and gold appears to be "dead in its tracks." I don’t have smoking gun proof for this thesis—yet—but empirical data are highly supportive of the notion.

The numbers above are aggregates for Bank Holding Companies in the U.S.—which fall under the purview of the Federal Reserve—which can be found in the quarterly derivatives report (see Appendix: Supplementary Graphs and Tables, Table 2).

Understand that the "Quarterly Derivatives Report" is published by the OCC. All but one chart/table in the OCC’s quarterly report (typically running close to 49 pages in length) details notional derivatives aggregates of Commercial Banks only. 

On closer examination, you might be wondering why ample information is provided for Commercial Banks, while only one page is published outlining the derivatives activity for Bank Holding Companies.

Here’s the difference: It’s all about who regulates. Commercial Banks are regulated by the OCC, while Bank Holding Companies in the U.S. are regulated directly by the Federal Reserve.

The material difference in quarterly derivatives reporting between Commercial Banks and Bank Holding Companies is that the Fed provides VERY LITTLE breakdown in the positions held by Bank Holding Companies under their purview. For instance, the OCC gives us a breakdown on Commercial Banks’ precious metals futures positions—but there is NO MENTION of precious metals, at all, in the limited reporting we get on Bank Holding Companies.

So, you may ask, why is this important or a "material shortcoming"?

Reason: Institutions like Morgan Stanley (MS) and Bank of America (BofA) trade almost exclusively in the name of their Bank Holding Company. If you take a peek at the total aggregate position of BofA, you will notice that in Q3/20 their total derivatives book size for the Commercial Bank is reported as 17.5(T), while BofA Bank Holding Co. reports a total derivatives book size of 34.4(T). The different notional amounts being reported by BofA are largely a "legacy issue" involving their absorption of Merrill Lynch as a Holding Company during the last financial crisis.

In pointing this out, I would like to share a personal anecdote: Back in early 2009, I had lunch with a very good friend who happened to hold a very senior trading position in the Treasury at Morgan Stanley, New York. Over lunch, we talked about "markets" and the size of derivatives positions at some of the large American banks. I told him that JPMorgan and Citibank—according to OCC reports—seemed to be the big players in the precious metals markets. He responded to me, "What about us (Morgan Stanley)"? He then went on to tell me that he personally knew the gold trader at Morgan Stanley New York—very well—and that Morgan Stanley New York was in the top two or three gold trading houses in the world.

I share this personal story to draw attention to the fact that the public gets very little detailed information regarding the tens of Trillions in notional derivatives being traded by Bank Holding Companies. I consider this INTENTIONAL MISDIRECTION on the part of regulators, namely the Fed—is anyone really surprised?

So, the reality is this: We get very imperfect information regarding outstanding notional amounts of derivatives held by American banking institutions—and with everything we do get being reported three months in arrears, giving an accurate account of "where we are" is akin to driving a car looking out the rear window making notes on the changes in scenery.

Each quarterly derivatives report begins with a "preface" very much like the one I have cut and pasted below from the Q3/20 report:

About This Report

The Office of the Comptroller of the Currency’s (OCC) quarterly report on bank trading and derivatives activities is based on call report information provided by all insured U.S. commercial banks and savings associations; reports filed by U.S. financial holding companies; and other published data. A total of 1,371 insured U.S. commercial banks and savings associations reported trading and derivatives activities at the end of the third quarter of 2020. A small group of large financial institutions continues to dominate trading and derivatives activity in the U.S. commercial banking system. During the third quarter of 2020, four large commercial banks represented 87.3 percent of the total banking industry notional amounts and 77.7 percent of industry net current credit exposure (NCCE) (see tables 1 and 4 in the appendix).

The reports also include an executive summary like the following from the Q3/20 report:

Executive Summary

  • Insured U.S. commercial banks and savings associations (collectively, banks) reported trading revenue of $9.0 billion in the third quarter of 2020, $5.7 billion less (38.9 percent) than in the previous quarter and $1.7 billion more (23.5 percent) than a year earlier (see table 1).
  • Credit exposure from derivatives decreased in the third quarter of 2020 compared with the second quarter of 2020. NCCE decreased $21.0 billion, or 4.1 percent, to $490.0 billion (see table 5).
  • Derivative notional amounts decreased in the third quarter of 2020 by $992.0 billion, or 0.6 percent, to $178.6 trillion (see table 10).
  • Derivative contracts remained concentrated in interest rate products, which totaled $129.8 trillion or 72.7 percent of total derivative notional amounts (see table 10).

Again, I want to point out that the report is (purposely?) disingenuous on its face: They tell us that four banks have/control 87.3% of the action. The reality (if you include 30+ Trillion from MS and another 17 Trillion from BofA) is that five banks control about 98% of the action. When five players control 98% of ANYTHING, you don’t really have a "market"—AT ALL. This is a CLUB—and we aren’t in it.

Addendum to 2020 Derivatives Wrap Up

Added March 25th, 2021

It is now late March 2021 and the OCC has released its Q4/20 Quarterly Derivatives Report.

I can now report that aggregate notional amounts for U.S. Bank Holding Companies have decreased (see Table 2 in Appendix: Supplementary Graphs and Tables) to 218 Trillion (down from 239.5 Trillion at Sept. 30/20), a drop of 21.5 Trillion. I want to impress upon everyone that these aggregate amounts are overwhelmingly comprised of interest rate contracts that have cash U.S. government securities (bonds and/or T-bills) embedded in them. Ergo, a dramatic drop in aggregates represents a REDUCTION in transaction demand for U.S. government securities. Correspondingly, we saw the yield on the 10-year U.S. government bond “back up” to just under 1.00% in Q4/20 from just under 75 basis points at Sept. 30/20.

The 4th quarter of 2020 was also extremely noteworthy for another reason: We witnessed a “complete decoupling” of what I broadly refer to as “the anti dollar trades,” namely, price differentials between precious metals and cryptocurrencies. At the end of August 2020, the price of Bitcoin was roughly 9,000.00 and the price of gold was 1,900.00 per ounce. Today, Bitcoin is above 50,000.00, and gold is struggling to “gain traction” at 1,740.00. I point this out because it appears to me that decisions were made back in Sept./Oct. 2020: Crypto prices were allowed to rise while precious metals prices have undergone some of the most aggressive nominal price suppression ever, despite there being record physical demand. Note that Oct. 1 in the U.S. is the beginning of a new fiscal year for the Federal Government (when new programs/ideas are typically rolled out) and the 10-year yield clearly vectored UP at that precise point (Oct 1/20). Given that rates have continued to rise substantially, my expectation is that we will see a further drop in aggregate notional derivatives when Q1/21 data are made available in late June 2021.