Trading Treasury Rates with ETFs - Part 2

In Trading Treasury Rates Part 1 I established a good reason for devising ways to trade U.S. Treasury interest rates without using the futures markets. In that article I offered two candidate ETFs, the iShares Barclays 20+ Year Treasury Bond Fund (TLT) and the Proshares Ultrashort 20+ Year Treasury Fund (TBT).

TLT is a long bond fund - it's value rises when the market value of long-term U.S. Treasury securities (USTS) rise (which happens when their market yields fall).

TBT in contrast is a short fund which does the opposite - it's value rises when the market value of long-term USTS falls. Further, unlike TLT, TBT is a leveraged fund and it is also a delta fund. A delta fund is designed to track the change in the value of an underlying security (rather than track the long-run value of the fund over time) day by day, percent by percent. If the fund is 2X short-leveraged like TBT, if on any given day TLT were to rise by X%, then TBT should fall by about 2X%.

The benchmark for both funds is the Barclays Capital 20+ Year U.S. Treasury Index, which is discussed in the next part of this series. TLT tries to follow the index exactly. TBT attempts to track two times (2X) the inverse of the daily change in this index each day.

Criteria for Choosing the ETFs

Now let me justify these two choices.

Of the many ETF candidates that we might choose, the following criteria should be satisfied

1. The ETF must be liquid, which implies that it should have high daily volume, especially when compared to peers.
2. If the strategy requires options, as our does, then a full range of put and call options must be available for the ETF, and the options that are near the money in near-term months must have both adequate open interest and daily volume.
3. The ETF should track well - meaning that if it promises to track an underlying index or other stated value, it should have a track record of doing that with a high degree of accuracy (the criteria would vary from one context to another, but in this context for a straight tracking ETF like TLT it should be above 95% and for a delta ETF it should be above 90%).

There are more than a dozen ETFs that track U.S. Treasury securities of various maturities. Candidates can be found in online searches at dedicated sites or in the Online Wall Street Journal Market Data section.

Daily Volume

Figure 2.1 U.S. Treasury Security Exchange Traded Funds (ETFs) shows a selection of mid- to long-term funds for both longs and shorts. A quick glance shows that only TLT and TBT have the necessary liquidity, as measured by average daily volume, to qualify for the first criterion above. The daily volume eclipses that of the other ETFs. The Proshares TBF may seem like a better candidate than TBT for short trades because it is not leveraged, but at 238,000 shares a day and no listed options it must be ruled out (N means no options, I means illiquid - inadequate open interest and/or daily volume).

Adequate Options

In Part 1 when describing general strategy I stated that if we were to use options for our strategy, we would use only puts and calls on TLT. If not using options then TBT provides a direct short opportunity but if options are required then puts on TLT are more suitable than calls on TBT.

Although the argument won't be supported with data here, because delta ETFs don't track with anywhere near the integrity of a straight tracking ETF like TLT, options on delta ETFs compound the error in tracking integrity. In a word, they are too volatile.

Even the TLT options barely qualify for serious trading. They tend to be liquid with adequate open interest and daily volume only right around the money. Also there is almost always a minimum spread between best bid and best ask of five cents and it is often higher. This is a large negative, because on many of the equity index ETFs this spread is seldom more than a penny.

Figure 2.1 Near-the-Money TLT Options shows the daily volume and open interest on TLT for May 20, 2010. As can be seen both volume and open interest is adequate for small trades (say 30 contracts or so) but not really for large trades. Although not shown here, it should be noted that volume and open interest on TBT options are at least as high as TLT and in volatile markets sometimes higher.

The Tracking Record

Traditional ETFs that track a known index over the long-run, like TLT, normally have a good tracking record (at above 95%) if they are popular and liquid.

Delta ETFs, which attempt to track only the change in the target day-by-day, sometimes with leverage like 2X for TBT, don't have such a good track record. If the Delta ETF has any sizeable error on the day-by-day target, that error can accumulate over time to where the long-term performance of the target has little correlation to the long-term performance of the ETF that tracks it.

This is partly because of how these ETFs are collateralized, which is discussed in the next part.

The criteria for measuring how well an ETF tracks its target, in this case the Barclay's Capital 20+ Year U.S. Treasury Index, compares the daily Net Asset Value (NAV) of the fund to that of the underlying index. ETF fact sheets and prospectuses can always be found on their parent web sites with a simple search and the fund performance relative to the target they track can be found in either or both.[1] If untrusting, a researcher with some spare time can use regression analysis to see how they track.

TLT tracks very closely, with less than 2% error because (a) iShares collateralizes TLT by simply buying the same securities as are used in the index in the same proportions, and (b) the stock is popular and liquid enough for the daily trading price to actually track the true Net Asset Value (NAV) of the holdings.

It turns out that even though TBT is a delta ETF and is collateralized with derivatives called Treasury Swaps, because it is also liquid, in the first quarter of 2010 only on a few occasions did it track with more than a 3% error. For short term trades this is completely adequate, although because this is a delta ETF, its usefulness as a long-term short hedge (or hedge against rising interest rates) is still questionable.

NEXT: Collateralization and pricing conventions for these two ETFs.

---

[1] In this case see the ProShares UltraShort 20+ Year Treasury Fact Sheet and Barclays 20+ Year Treasury Bond Fund (TLT) and once on that page, link from Related Links and Documents to the "Tracking Error Chart."

Trading Treasury Rates with ETFs - Part 1

Suppose you wanted to trade interest rates on long-term U.S. Treasury Securities, either as a hedge or a speculative position.[1]

For example, suppose you believe, as do I, that inflation will emerge as a serious problem in the next few years because of our debt excesses. If that happens, interest rates will rise on bonds in general, including U.S. Treasury notes and bonds and, for those bonds already issued before the inflation, their market values will plunge, causing large capital losses for holders. Also the stock market typically performs poorly during periods of high inflation. Therefore a trader with fixed income investments and equities might desire a strong hedge against inflation and rising interest rates.

The futures markets offer a huge selection of hedges against interest rate movements. But not all traders like to trade futures because they dislike the implied leverage and relatively high cash requirements for trading futures, or they simply feel uncomfortable in that arena.

Exchange Traded Funds (ETFs) now offer equity traders, including those operating on a very small scale with limited budgets, a full range of strategies for playing interest rates of different markets, from U.S. Treasury Securities (USTS) to commercial junk bonds, of all maturities, from less than one year to thirty years.

This series of articles will discuss trading with interest rate strategies for long-term USTS primarily using two common and liquid ETFs, iShares Barclays 20+ Year Treasury Bond Fund (TLT) and the Proshares Ultrashort 20+ Year Treasury Fund (TBT) and put and call options for TLT.

The General Strategy

Generally the strategy will be quite simple: (1) If you think that interest rates will fall on long-term USTS then you will buy TLT or buy calls on TLT or (2) if you think interest rates will rise on long-term USTS then you will buy TBT or buy puts on TLT. (I could have included options on TBT in this mix but didn't for reasons discussed in a later article).

The general strategy is easy to explain, but once we consider questions about the proper size of trades, trade timing, and the relationship between interest rate activity on long-term USTS and the market price of USTS and, hence, these two ETFs then the specific strategy becomes much more complicated.

So we will explore this in steps. The remainder of this article is dedicated to describing and understanding the mathematical relationship between the market yields and market values of USTS.

Future articles in this series will explain (1) the TLT and TBT ETFs and how they are structured and why they are suitable, (2) how TLT is securitized and how that affects TLT's price, (3) how TLT puts and calls respond to TLT volatility and interest rate volatility, and (4) the development of trading models that make use of all of this information.

The Relationship Between Interest Rates and Market Values of USTS.

For all yield-bearing (interest paying) financial assets (YBFAs) traded in the secondary markets after their original issue, including U.S. Treasury Securities of all durations, their market values move in opposite direction of their market yields.

For example, consider a 30-year bond originally issued with a par value of 100 paying a coupon rate (original issue rate) of 4% of par annually. If 10 years later newly-issued 20-year bonds were yielding 6% at par, then the 30-year bond with only 20 years remaining in its life would have to also yield 6% to remain competitive. Because the coupon rate is a fixed constant throughout the life of a bond, the only way the old bond can raise its yield is through a reduction of its market value below par. The bond would fall to a value below par and in this case would in fact fall to a value of $77.60.

Where did I get that number?

The relationship between a bond's market value and its market yield is mathematical. Figure 1 Simple Bond Valuation Formula shows that mathematical relationship, and provided the solution for our bond problem above.[2]

Figure 2 Possible Values for 30-year Bond taken from a slide used in a lecture to explain this relationship, shows all of the possible values that this 10-year-old 30-year-bond would assume given a range of possible yields offered by newly issued competitive 20-year bonds. As can be seen in the ranges shown, the old bond can trade as high as 155 (when the new market yield is 1%) and as low as 34 (when the market yield is 14%).

It should be easy to see that in an environment of volatile interest rates a long-term bond like this example can have the same degree of price volatility as some stocks. More important, in an inflationary environment, if long-term bond yields rise to reflect the inflation, then long-term bonds will suffer serious capital losses.

Consider a real example. As we will see, one of the USTS assets used the collateralize TLT is a 30-year Treasury bond that mature on November 15, 2039, has a coupon yield of 4.38%, but on April 30, 2010 had a market yield of 4.53%, slightly higher, and therefore was trading at a discount price of 97.52. Suppose that by November 15, 2013 that because of double-digit inflation this bond had a market yield of 12%. In that case this bond would be trading for 39.84!

So in an emerging inflationary environment, you would want to be short in this bond, or short in something that tracks this bond.

NEXT: The tracking stocks (ETFs) and their options that we will use to do this.

---

[1] If the reader is unfamiliar with the types of U.S. Treasury Securities available on the secondary market, review The Market for U.S. Treasury Securities.

[2] For the curious reader, this mathematical relationship and why it exists is explored in much greater detail in the document Bond and Note Valuation and Related Interest Rate Formulas. The formula shown in Figure 1 is derived in this document and more complicated bond formulas are also developed and explained.