Global Derivatives Watch

Tracking trends in the derivatives (swaps, options and futures) markets is key to understanding how investors and traders hedge their portfolios and positions. Like debt (bonds) and equities, derivatives are also subject to risks, such as interest rate risk, default and counterparty risk, liquidity risk, market risk (volatility), futures curve risk, etc. In particular, it is relevant to study the stability of the interest rate swaps market with respect to volatility in interest rates (interest rate risk) and the stability of the credit default swaps market with respect to default or counterparty risk. The growing size of these markets reinforces the need to track and study market stability.  

There are several key data sources for market data on derivatives, the Bank of International Settlements (BIS), TriOptima (interest rate derivatives), and the Depository Trust and Clearing Corp (DTCC) (credit default swaps). Below are initial data compiled from these sources, namely the outstanding size of derivatives market categories over several time periods. As time progresses, I plan to add more metric detail from these and other sources depending on the significance toward a particular analysis or study. ISDA, the International Swaps and Derivatives Association, publishes an expanded data source list HERE.

On “notional amounts outstanding” – In general this means the value of a derivative’s underlying assets at the spot price, the face value that is used to calculate payments made on a derivative, or the total value of a leveraged position’s assets. More specifically, in the case of swaps this refers to the notional principal amounts upon which the cash flow payments are calculated, which determine the value of the swap; in the case of futures (forwards), this refers to the unleveraged total value of a leveraged contract, e.g. L•$X for a $X contract that is leveraged L times; in the case of options, it is the total value of the shares or contracts controlled by the option, which is leveraged $X/$Y, where $Y is the price paid for the option.

ISDA has a point when they state “notional principal amounts overstate exposure as they do not reflect the market value of the underlying contracts and benefits of close-out netting.” The BIS data below (and my comments) acknowledge the market values with reference to notional amounts as a focus. “Close-out netting” is defined as the termination of obligations under a contract with a defaulting party and subsequent combining of positive and negative replacement values into a single net payable or receivable. Since not all derivatives contracts (especially CDS contracts) are necessarily handled that way, there remains an issue when a contract is insufficiently collateralized or when there exists insufficient capital to make good on an obligation following a default. This was especially the case with AIG and others during the height of the 2008 financial crisis, whereby the underlying CDO instruments lost significant value and AIG was severely undercapitalized to pay out on the CDS contracts it had sold to cover CDO losses. Unfortunately, data to show close-out netting and collateralization does not really exist on a sufficient level to gauge true credit exposure, though ISDA estimates this exposure “on average.” (See more below under ISDA data.) In my view, more granularity on credit exposure is of value to the markets, and that means more detail on close-out netting, collateralization, and capital reserves.

Notes on the latest data and trends (8/29/2011):

9/7/2011: Added DTCC market share data and central clearing comments; Added ISDA data.
9/12/2011: Added ISDA and BIS historical data.

BIS data: The largest derivatives market by far is the interest rate swaps market, which in Dec 2010 commanded ~61% in terms of both notional value and gross market value. The steady growth of the forex (16%), currency (29%) and interest rate (7%) swaps markets from 2008 to 2010 suggests an increasing demand to hedge against currency and interest rate risk exposure. The commensurate reduction in gross market values of currency and interest rate swaps contracts reflects an overall decrease in interest rates upon which cash flows are calculated; the commensurate reduction in market values for forwards and options would imply greater leverage. The commodities futures and options markets contracted by over 30% from 2008 to 2010, perhaps reflecting the generally slower economic growth and less of a need over that period to hedge against higher commodity prices. Future data in that category may show the opposite, as parties seek to hedge against commodity inflation. (The tell is that the gold futures markets remained stable over the same period.) The contracting size by ~29% of the credit default swaps market reflects a decreased demand for protection against default risk, but demand may increase in the future as parties seek to hedge against increasing sovereign debt defaults and linked financial institution defaults. BIS historical data, integrated with ISDA historical data, shows the exponential growth of the derivatives markets since the late 1980s, and is best viewed on a logarithmic scale to compare notional amounts and gross market exposures - this is shown below. 

TriOptima data: This data shows the latest size of the interest rate derivatives market in category detail. Historical data from TriOptima is located HERE. The data lines up approximately with BIS reports, which are only available on a delayed quarterly basis. This latest TriOptima data implies that the market size has grown by ~15% in just the first 7 months of 2011 alone. Of note, 50% of the contracts are cleared through a central clearinghouse, while the remainder is cleared through OTC dealers. Such central clearing has been increasing steadily, mainly taken from “G14” OTC dealers. If we are to believe that central clearing decreases counterparty risk, this is a positive trend. TriOptima now breaks out the overnight indexed swap (OIS), where the cash flows of the floating rate leg of the swap are calculated using an overnight reference rate compounded over every day of the payment period. That overnight rate, or OIS rate, varies depending on reference rate; in the US the rate is the overnight Federal Funds rate. The contrast is the LIBOR rate, upon which many plain vanilla interest rate swaps are based. In the past, LIBOR has shown extreme volatility due to sudden tightening in interbank lending from increasing bank counterparty risk and other factors, such as liquidity issues. The growth in the OIS category reflects a rising demand for swap lines based on relatively stable rates in the short-term funding markets. From Dec 2010 to Aug 2011 the share gained over 27% in terms of notional value. (The LIBOR-OIS spread is a key metric for gauging the health of global credit markets.)

DTCC data: This data shows the latest size of the single-name and multi-name credit default swaps (CDS) market in category detail of open positions. Historical data from DTCC is located HERE. The largest demand for single-name CDS contracts is on behalf of the financials (~22%) and sovereign government (~18%) sectors. From Jan 2011 to Aug 2011 demand in terms of notional value has gone from $14.45T to $15.83T, with the largest increase from the sovereign government sector ($2.5T to $2.9T). Multi-name contracts involve CDS indexes and tranches of indexes, particularly those linked to mortgage or other asset backed securities (ABX), credit entities in North America and emerging markets (CDX), and credit entities in Europe and Asia (iTraxx). The largest outstanding position is in the iTraxx Europe (~31%), perhaps reflecting the hedges against defaults stemming from the European sovereign debt crisis. The second largest position is in the CDX NA Investment Grade (~21%), perhaps reflecting the demand for protection against investment grade debt/bond degradation. From Jan 2011 to Aug 2011 demand in terms of notional value has gone from $10.88T to $13.63T, with the hedges against the Europe exposure having changed little in percentage terms. The total notional value size of the single- and multiple-name (index-linked) CDS markets reported went from $25.33T to $29.46T from Jan-Aug 2011. Of the latter, approximately 89% of the contracts were negotiated through a dealer, while 11% were privately negotiated between a non-dealer and the customer. What percentage of these were cleared through the four clearinghouses that currently handle CDS clearing (ICE, CME, Eurex, LCH.Cleannet) is not readily available through DTCC in their dataset, though this is certainly pertinent information. Generally clearinghouses would handle standard plain vanilla CDS contracts, and not the more exotic, specialized, and one-off CDS contracts, though the latter historically have been at the center of the controversy regarding high risk and undercapitalization. Arguments have been made for/against central clearing and exchanges for OTC CDSs (see HERE, HERE and HERE), and the only way to really make the case that central clearing features (contract netting and collateralization) and the price transparency and liquidity offered by an exchange both contribute an economic value to the market is to see that participants are choosing such a route in greater proportions and that market data show risk mitigation benefits. Interest rate swaps have a longer and more established central clearing history, as the TriOptima data indicates, and ISDA through its end-user surveys and research studies has persuasively shown the impact of clearing, netting and collateral on notional amounts and risk exposures.

ISDA data: ISDA market and end-user surveys are located HERE. ISDA publishes historical data on derivative market size as measured by notional amounts outstanding via market surveys going back to 1987. The table “historical derivatives market data” below reflects the compounded growth in the IR/Currency, CDS and Equity derivatives markets, plus indicates the height of these markets during the pre-bust 2H ’08. Further, according to ISDA, the total derivatives market size in terms of notional amounts outstanding is overestimated by BIS due to contract clearing (mainly of interest rate instruments): BIS counts two contracts outstanding that are really structured as one through the net clearing level. ISDA”s modifications to BIS data for these factors are shown in Table 1-ISDA and 2-ISDA at the bottom. From this modified data, ISDA estimates on average the risk mitigation benefits (gross credit exposure) of netting and collateral in Table 3-ISDA. As mentioned above in my general commentary, this is an average estimate and says nothing about the existence of highly leveraged contracts that are not netted and do not have sufficient collateral or capital to cover large losses, should there be tail risk failures of the underlying instruments (this is especially true of CDS, but also of IRS). 

–IRS/TriOptima Data:


–ISDA Data: