Taxable PIMCO CEFs remain one of the best-performing and hence best-regarded funds in the world of income investing. There are many drivers behind the stellar track record: elevated leverage, exotic holdings, the use of auction-rate preferreds, at-the-market offerings and others. In this article we take a look at one of the less well understood drivers of this outperformance – interest rate swaps. We use the Income Strategy Fund II (PFN) as a case study.
The ideas in this article grew out of a long discussion with a reader who has been banging the loudest drum in the investor community about PIMCO’s use of swaps in their taxable CEFs. The two main takeaways are that swaps are significant contributor of income to taxable funds and that PIMCO uses swaps to transfer current income to future income via occasional modifications of the swap portfolio and its change in market value.
A Roller-coaster Ride On Distribution Coverage
PIMCO provide the most amount of disclosure about the workings of their CEFs. More disclosure is by definition better than less disclosure, particularly for more hands-on income investors. However, because this is not the same as complete disclosure, it is a case where “a little knowledge can be a dangerous thing”. To illustrate consider the following tongue-in-cheek chart that plots 6-month rolling distribution coverage for PFN – one of the PIMCO multi-sector CEFs.
Source: Systematic Income, PIMCO
Investors who diligently follow the monthly releases are likely experiencing whiplash with each release. Normally, distribution coverage tends to follow a smooth and fairly stable curve. This is for two main reasons: first, fund turnover tends to be moderate and even during periods when the pace of turnover is elevated the portfolio income production levels are not hugely impacted since assets leaving the portfolio don’t look completely different than assets entering the portfolio. And secondly, distribution coverage is, by definition, a rolling average which smooths out income generation across several months.
None of this really applies to PIMCO multi-sector CEFs which have distribution coverage that is not readily explained by the typical drivers of earnings. Unlike PIMCO municipal funds, multi-sector funds have distribution coverage that behaves wildly. The chart below shows the coverage of the multi-sector CEF Income Opportunity Fund (PKO) versus a more well-behaved PIMCO municipal CEF that boasts a more “normal-looking” coverage profile.
Source: Systematic Income, PIMCO
This is not to say that analyzing the traditional drivers of fund earnings such as leverage changes and sensitivity to short-term rates is completely useless. Indeed knowing the fact that PKO has added the most borrowings since the drawdown was a valuable piece of information.
Source: Systematic Income, PIMCO
It goes a long way in explaining why this fund has been able to outperform all other multi-sector CEFs in growing coverage since then.
Source: Systematic Income, PIMCO
However, what this analysis doesn’t give you, at least not for PIMCO CEFs, is a precise explanation of the month-to-month gyrations in coverage. And by extension it makes it difficult to calculate the “correct” absolute distribution coverage of the PIMCO CEFs.
To be clear we don’t need to know absolute levels of distribution coverage in order to position across PIMCO CEFs. Knowing the relative earnings path of funds is often enough. If we know that PKO will grow its earnings relative to other funds we can tilt to PKO and avoid funds like RCS and PGP which ended their deleveraging campaign by cutting their distribution, something we suggested may happen a few weeks prior to the fact.
However, for many investors the question of absolute coverage is important. This is because they want to know the true earnings yields of these funds rather than their relative earnings strength. This helps drive the decision whether to allocate capital to PIMCO CEFs or elsewhere. A fund that has a current yield of 10% but an earnings yield of 7% is much less appealing, particularly, since many other CEFs in the market earn roughly the same level without the sky-high premiums of many PIMCO CEFs.
A Look at PFN Net Investment Income
The chart below shows the year-to-date net investment income of PFN starting from the start of its last fiscal year in August of 2019. Overall this chart makes a lot of sense. Broadly speaking the line goes up in somewhat of a straight line for the duration of the year. However, there are a couple of weird months, highlighted in red, where the line basically pauses for a month.
Source: Systematic Income, PIMCO
A clearer chart which shows the monthly marginal income is shown below. In a “typical” month, the fund delivers something between 4 and 12 cents of income per month. The actual monthly income will vary because of two main factors. First, the fund holds securities with different payment conventions. Floating-rate securitized assets typically pay on a quarterly basis while bonds tend to pay on a semi-annual basis. And secondly, the actual securities will have different schedules – some bonds will pay on January and July while others will pay in March and September and so on. These differences create a normal variation in coverage.
Source: Systematic Income, PIMCO
Calculating coverage on a six-month rolling basis will compensate for these variations but it will not compensate for three other factors. First, floating-rates will vary on a daily basis and affect coupons of floating-rate securities such as loans and MBS. Secondly, turnover will ensure that the portfolio holds securities with different coupons even if they are the same in risk terms to the securities they replaced. The last factor is leverage which scales fund earnings up or down relative to net assets. PFN borrowings, shown in the chart below, have moved around quite a bit since the start of the year and would have had an impact on the fund’s earnings.
Source: Systematic Income, PIMCO
The trouble is that while this information is very useful to investors it doesn’t actually explain what happened to PFN income over these two months. So, what went on in October-2019 and June-2020 – the two months where the fund earned nothing. Turnover and changes in short-term rates are too small to make such an impact and leverage actually grew sharply in June. Unless the fund decided to move into Treasury bills during these two months, these explanations are not sufficient to understand the drivers of income.
Meet Interest Rate Swaps
Interest rate swaps are derivatives that create an exchange between floating-rate and fixed-rate cashflows and, consequently, allow a party to manage its interest rate exposure and duration. PIMCO uses interest rate swaps for three main purposes.
First, they are used to take a view on interest rates and the yield curve, in particular. Historically, PIMCO have maintained a yield-curve steepener position in their taxable CEFs. This allows the funds to make money when the yield curve steepens or when the differential between short and long-term rates widens. Interest rate sensitivity is often measured by something called DV01 which stands for a change in dollar value for a one basis point move in interest rates. The chart below shows the official April release from PFN for their overall portfolio sensitivity to changes in interest rates. The chart shows that the fund makes about $280k for every 0.01% increase in rates at around the 30-year point and makes about $450k for every 0.01% decrease in interest rates below the 30-year point, all else equal.
Source: Systematic Income, PIMCO
PIMCO have often commented on their curve steepening view. For example, in 2018 the fund company’s view was the curve rarely inverts and that longer-term bonds did not sufficiently compensate investors for the risk of inflation. This view was less prescient than their concerns about the end of the interest-rate cycle and the potential for bond supply to increase sharply.
The second key way in which the swaps portfolio changes the fund’s risk exposure is by reducing its overall interest rate sensitivity. Without the swaps the portfolio DV01 would be something on the order of $400k meaning the portfolio would gain $400k for each 0.01% reduction in interest rates across the curve, all else equal. The fund’s actual DV01 is on the order of $200k which lines up well with the fact that its swaps portfolio DV01 is around $200k paid rates. What this all means is that the fund has halved its sensitivity to parallel changes in interest rates. There are different ways to think of this. Traders have an insider joke to call a winning bet an active choice on their part and a losing bet a “hedge”. As an active choice the positioning in swaps clearly did not work out this year given that rates have fallen sharply across the curve. However, the swaps can also viewed as a hedge that reduces the fund’s exposure. If interest rates went up sharply, particularly in the long end, the fund would have seen its NAV fall less, all else equal, than those funds that did not reduce their interest rate sensitivity.
The second use of swaps to manage the funds’ exposure to floating-rate cashflows. For example, as of April-2020 PFN had $100-150m of its $720m portfolio in the form of floating-rate assets such as loans and RMBS. There is some uncertainty because not all RMBS are floating-rate and PIMCO does not disclose further details. The loan part of the portfolio, at around $109m notional, is composed of assets with maturities of 3-7 years. This corresponds reasonably well with the $150m or so of LIBOR-pay swaps held by the fund since RMBS and ABS will also contribute to the floating-rate asset base somewhat.
Source: Systematic Income, PIMCO
Over the past 6-months or so this has worked our very well for PIMCO given that the loan CEF sector has seen the highest number of distribution cuts in the last few months relative to the sector size all mostly due to the sharp drop in LIBOR.
Source: Systematic Income, PIMCO
Thirdly, and perhaps most importantly, PIMCO uses swaps to boost fund income. This strategy itself is composed of three elements.
The first element is that, as a base case, the steepener position of the swap portfolio generates income. This is particularly true in a flat yield curve environment because you have to pay fixed on a much smaller notional since the duration of the longer term swap is much higher. For example, given the last PFN published portfolio if we set all swap coupons to the same figure the portfolio would generate $1.5m a year or about 0.27% in additional yield on net assets.
The second element of this strategy is that often PIMCO will turn some longer-dated paid-fixed swaps into forward-starting contracts. These swaps have tended to be negative carry for PIMCO given LIBOR has tended to be at depressed levels outside of the 2018-2020 episode. By making them forward-starting PIMCO, in effect, shuts off cashflows in these swaps but continues to benefit from their interest-rate exposure profile.
For example in its October-2019 portfolio PIMCO had all of its long-dated paid-fixed swaps as forward-starting (marked with a footnote and highlighted below).
This meant that $290m of its $435m negative carry swap portfolio had cashflows switched off. In terms of the impact on the portfolio income profile – had all swaps been spot-starting PIMCO would have received $5.6m over the next year. However, with some swaps being forward-starting that increased to $9.9m over the next year. In yield terms that was an increase from 0.88% to 1.56%.
The third element of this strategy is that PIMCO artificially lower the coupon on the longer-dated paid-fixed swaps. The swap holdings as of Apr-2020 are shown below with the pay-fixed swaps highlighted. These swaps have coupons well below those of receive-fixed swaps. When PIMCO initiates such off-market swaps we can usually tell from the fact that they are executed with a significant premium paid by PIMCO which reflects the fact the rate it pays is below market.
The last time the fund adjusted pay-fixed swaps to a lower coupon was between July-19 and October-9 (we don’t have data past April-2020). The fund restructured a $283m 2.5% 2048 payer swap into 5 2050 payer swaps with a weighted average coupon of 1.71% and a total notional of $272m.
In this particular case PIMCO restructured the swaps in a way as to pay 0.79% less on about $280m for 30 years. In present-value terms we estimate that to be around $50m. So PIMCO would have had to compensate its counterparty by that amount. This would have caused a ripple effect in two other ways. First, PIMCO would have had to economically transfer this amount to the counterparty. Secondly, the market value of the swaps would have had to increase by the same amount.
We can see the impact of this second point on the market value of the swaps which grew from $34m to $94m from July to October-2019. Part of this was, of course, changes in interest rates across the curve but the bulk of it was this change.
On the first point it looks like this restructuring caused a drop in both NII and UNII. What was the impact across these metrics? In order to answer this question we need to know how much income PFN tends to generate on a “normal” month, that is on a month where it does not mess with its swaps book. And secondly, we need to look at how much income and UNII it reported in October versus the previous month.
If we look at the six months where the swaps book was untouched between October-2019 and April-2020 we see that PFN generated on average earnings per share of $0.0683. In October 2019 the fund generated no income and the fund’s UNII also dropped by $0.20. If we sum the two together multiplied by the estimated number of shares on that month we get to missing income of about $18.5m. This figure is well short of the $50m of capital necessary to restructure the swaps so this remains an outstanding question.
This illustrates an important point in how PIMCO manage their taxable funds using swaps. By restructuring the swaps PIMCO are able to move capital from current income to future income via a change in swap market value. The required capital transfer for the swap restructuring came out, in part, from NII and UNII but it also increased unrealized gains and hence future income. By lowering the coupon on the swaps PIMCO will be able to enjoy a higher annuity stream in the future. Ironically, the drop in current NII was in this case actually a sign of stronger income in the future rather than less. What do we expect the impact of this to be? Relative to current net assets the additional annuity stream is around 0.37% in yield terms.
Is this an efficient use of capital? On the one hand, not really. The fund could have instead bought a 6% yielding security to produce an income stream of about 0.50% over the next few years (= $50m x 6% / $596m). However, a big difference here is that the income from the swap is effectively risk-free and it is there for 30 years whereas assets could default or see yields fall creating reinvestment risk for the principal.
Takeaways
Swaps are a significant and volatile contributor to income of PIMCO taxable CEFs. As of the last filing, swaps contributed about 1% of additional yield to PFN. And they are also a volatile contributor because over the last year this contribution ranged from 0.6% to 1.7% in yield terms relative to current net assets. This is due to PIMCO’s penchant of tweaking the swaps portfolio.
Investors who pay attention to monthly changes in coverage and UNII have to take them with a grain of salt since they don’t provide the full picture of what happens inside the funds. Since PIMCO can move current income to future income via a change in swap market value, a drop in current income and UNII is not necessarily an indicator of lower earning capacity going forward.
To get a cleaner picture of the funds’ earnings and hence distribution capacity investors should do two things: they should select periods when swaps are left alone (such as the period from October-2019 to April-2020) and they should be aware of how much income comes from the swaps since this part of the funds’ income is likely to be volatile. As far as PFN, over this period we calculate that the fund’s NII yield was around 9.72% based on the July-2020 net asset amount with 1.25% of this coming from the swaps portfolio, leaving 8.48% as the “organic” net investment yield of the underlying portfolio.
The low level of income we just saw in June of this year and the drop in UNII was likely another rebalancing of the swap portfolio rather than any indication of a sharp drop in earnings capacity. We should be able to confirm this once the next holdings portfolio is released in the coming months.
Check out more in-depth and timely commentary as well as Income Portfolios and interactive Investor Tools at our Premium service.
ADS Analytics LLC / Systematic Income provides opinions regarding securities and other related topics on an impersonal basis; therefore no consideration is made towards your individual financial circumstances.
All content presented here is not to be regarded as investment advice or constitute a client / advisor relationship. It is for general informational purpose only.
Trading securities involves risk, so you must always use your own best judgment when trading securities. You assume the entire cost and risk of any trading you choose to undertake. You are completely responsible for making any investment decisions.