One might think that treasury bills protected against inflation (TIPS) are precisely an appropriate investment in times of high inflation. How do these bonds work and what is the best way to invest in TIPS?
On September 11, I published “Amazing 9+% Yields In Inflation-Protected Treasuries(“previous post”). This sequel is prompted by the discussion that followed the post.
How do TIPS work?
In the previous post, my description was rather terse. Allow me to correct my mistake.
Most investors know that TIPS are government bonds currently offered in 5, 10 and 30 year maturities. Like all treasury bills, they carry no credit risk and are not taxed at the state level. The coupon rate is fixed at the time of issue, but the principal is adjusted according to the evolution of the consumer price index (CPI). However, the process of this adjustment is not so simple.
Among several types of government-issued CPIs, TIPS are adjusted based on the so-called non-seasonally adjusted consumer price index for all urban consumers (CPI-U) (you can check the last report of august here). The historical results of this index and its year-on-year variation are shown below, copied from a government site.
In October, the price of TIPS will be adjusted according to the results of August, ie with a lag of two months. The Treasury has already released the chart that the inflation factor will change daily in October.
The right column in the image below shows the inflation factor for the bond maturing on January 15, 2023 (CUSIP 912828UH1) that we highlighted in the previous post. Data for this bond and others are available on the government’s TreasuryDirect website.
The numbers in the table may be too small, but please take a closer look at the right column. Here’s something that might surprise you: the inflation factor is down slightly from October 1 to October 31 despite high annual inflation in August of 8.3%. How come?
From our first chart, the CPI numbers were 296.276 and 296.171 in July and August, accordingly. This minor drop in the unadjusted inflation number carries over to the inflation factor for all TIPS in October.
For a better understanding, let’s move on to real life quotes. The table below is copied from Fidelity on September 16. The two colored rows are the bonds recommended in the previous post (they jumped a day after I posted and are not so attractive anymore). The January 15 obligation ranks first. On September 16, its inflation factor was 1.28363 on Fidelity but according to our chart, it will only be 1.28357 on October 1.
From my readers, I know a lot of them didn’t know how inflation adjustment works.
The most important conclusion of this section is that even in times of high annual inflation, inflation adjustments can be negative for a limited period.
One of my readers suggested me a good site to track TIPS related issues that I was unaware of – tipswatch.com
TIPS market efficiency
In the previous post, I had highlighted an abnormally high YTM for the two ultra-short TIPS in yellow in the image above (they were 4.351% and 2.959% respectively when I submitted my post previous). Many comments were related to the efficiency of the market: how could it be when thousands of experts with powerful computers and sophisticated models monitor the Treasuries day and night?
To this I must reply that the inefficiency of the market was very limited, if at all. The attractiveness of ultra-short TIPS was linked to BOTH their high YTM and their expected negative correlation with stocks. In the worst case scenario for these TIPS (i.e. lower inflation by expiration than economists forecast on the day of the previous post), I expected their returns to be similar to those of regular Treasuries, but with decent expected returns from equities. Otherwise, I expected TIPS to beat regular Treasuries. A day after my message and contrary to economists’ expectations, the CPI figures turned out to be high. TIPS jumped but stocks fell. Regular Treasuries do not have this negative correlation, as they are intended to return their yield regardless of the CPI.
Also, market efficiency was not exactly what it seemed. Below is the depth of the trading portfolio for Fidelity’s January 15 bonds on September 16.
Please pay attention to the Qty column under “Asking Prices”. The best price of $99,068 is available for only an order of around $6 million. If you want to buy the title in significant quantities (~$500 million at the bottom row), the price becomes $99,500! It’s a sinkhole for ultra-short bonds.
On September 7, the depth of the January 15 book was similar and I bought a notable fraction of $6 million. After that, bond prices started to climb. Could my actions have slightly affected the market? It sounds extremely dubious, but it still haunts me.
TVIP, STIP and PBTP
Many investors have decided to allocate to TIPS since inflation became a hot topic in mid-2021. Using ETFs seems the most practical option for this purpose. Due to rising interest rates, ETFs that only buy short-term TIPS have become particularly popular. My previous post reviewed two well-known instruments, namely sponsored by Vanguard (NASDAQ: VTIP) and Black Rock (NYSEARCA:STIP) (they are very similar – please see previous post for minor differences).
This strategy of using ETFs was not successful in 2022 (it worked much better in 2021). A reader of mine who invested in STIP noticed the following – I’ll quote it with minor omissions:
…looking at my STIP for the past year, I see that I have a net decline of 1.1% (i.e. decline in principal > interest income). Therefore, cash would have been a preferred asset. Since I hold the STIP in a taxable account, I will pay taxes to the Fed on the interest income from the ETF. Therefore, my “real” return is not -1.1% but closer to -3% (after tax).
VTIP and STIP emulate the Bloomberg US 0-5 Year Treasury Inflation-Protected Securities Index and their average maturity (and duration) is ALWAYS close to 2.5. In reality, their duration is even higher: 2.6 for VTIP and 2.7 for STIP. This makes them sensitive to rising interest rates and explains what my reader experienced in 2022. Federal taxes exacerbate the situation. Ultrashort individual TIPS seem to be a better choice than VTIP or STIP for the next few months, as they have both a higher YTM and a shorter duration.
One of my readers asked me if there were better ETFs. The answer is yes – welcome to Invesco PureBeta 0-5 Yr US TIPS ETF (BAT:PBTP) with details available here.
It has a slightly higher expense ratio – 0.07% vs. 0.04/0.03% for VTIP/STIP. But it tries to mimic a different benchmark – ICE BofAML 0-5 Year US Inflation-Linked Treasury Index℠. For this reason, its duration is significantly shorter – 2.16, and it will be less sensitive to interest rate increases. Over the next few months, it is likely to perform better than VTIP/STIP. Due to its smaller YTM (1.18%), we still expect it to underperform ultrashort individual TIPS.
And the final observation: my readers have mentioned that some brokers (TD Ameritrade, for example) do not allow trading of individual TIPS. Most of them do, however, and one can find individual obligations based on their CUSIP. CUSIPs, in turn, can be found through various sources, including the TreasuryDirect site already mentioned.