Recently, I was asked what I thought was the biggest risk to the municipal bond market over the next 12 months. Interest rates rising? A credit event? Retail outflows? Supply? Or liquidity? I chose retail outflows. I think the biggest risk to the municipal market is more on the demand side than the supply side. Muni rates follow rates in general over the longer term; however, given it is a much smaller market and dominated by retail investors, technical factors, i.e. supply and demand, can greatly impact its relative performance over shorter time horizons.
Right now, I think the risk of outflows due to rate concerns is a bigger risk than a credit event (although this seems to change daily as new economic data is published). Most forecasts are not expecting dramatically higher rates, but changes in rate outlooks due to economic data or Fed actions/guidance could drive changes in asset allocations, which in turn, could trigger large retail outflows.
Credit events can also trigger outflows, particularly retail outflows. Because retail investors tend to be more reactionary to headlines, credit events similar to Detroit and Puerto Rico can trigger outflows. However, I think that the headline risk is lower than in recent years because even retail investors have come to see credit events as one off events not necessarily impacting the broader municipal market. Although our credit team worries a bit that we’ll start to see a lot of negative headlines due to the implementation of new pension disclosures for state and local governments. It remains to be seen if this will impact the broader market because it may not be as isolated as more recent credit events.
Given the relative size of the municipal market to other fixed income classes, retail outflows have a major impact on liquidity and therefore, municipal bond prices. This impact is exasperated by that fact that since 2008, the street no longer makes much of a liquidity commitment to the municipal market. Thus, outflows translate into heavy selling without the street taking inventory. Liquidity dries up very quickly and strains valuations as sellers are forced to take prices that reflect oversold conditions. Periods of illiquidity generally become a buying opportunity for buyers not constrained by retail flows, particularly crossover buyers. The oversold valuations eventually correct as crossover buyers come into the market to take advantage of the cheap muni yield ratios. This situation was best illustrated in December 2010 and early 2011 with Meredith Whitney’s infamous call for “hundreds of billions of dollars” in municipal bond defaults. Muni/treasury ratios rose to extraordinarily cheap levels as retail investors sold their muni bond funds. Eventually, institutional crossover buyers were attracted to the cheap valuations and bought munis; which meant that over time, the ratios reverted to more normal levels. It proved to be one of the most lucrative times to buy municipal bonds.
Of course, while I think the risk to municipals is more on the demand side, the supply side has risk as well. There are two types of municipal supply—refunding (refinancing at lower rates) and “new money.” In low rate environments, like we are now experiencing, refunding volume tends to increase. For instance, year to date we have seen increased volume relative to last year due to refunding transactions. If rates remain low there is some risk of increased supply that could weigh on municipal bonds. However, while that risk may result in more volume than last year, supply would still be moderate from a historical perspective. Also, while the increased supply could negatively impact municipal valuations, I believe the impact would be minimal as buyers could view periods of heavy supply as buying opportunities. I see new money volumes as continuing to be constrained by political factors; politicians are still very reluctant to explain increased debt levels to voters.
The good news regarding these technical factors is that while it produces risk, it also produces opportunity. And that’s why non traditional municipal buyers keep a careful eye on municipal ratios.
A month ago, I began my blog post with the following sentence: “Clearly, last Friday’s employment report increases the likelihood that the US Federal Reserve (Fed) raises rates at its June policy meeting, or soon thereafter.” Today, we have exactly the opposite situation. Literally, I could just replace “increases” with “decreases” and call it a day. However, a debate is simmering within our regular investment strategy meetings here, about whether it really matters if the Fed hikes in June, September, or December. If we take as a given that the Fed is ready to start a policy tightening cycle “soon” (for the first time in 11 years), who cares if it’s 11 years and zero months or 11 years and six months?
In short, it shouldn’t matter much but the market seems to care. Big swaths of interest rate products are repricing on every swing of the liftoff date pendulum. I touched on this in Americans Bearing Gifts, but just after the February employment report was released on March 6, the market appeared certain that the Fed would hike in June, and 10-year Treasury yields went to 2.25%. Fast forward to last Friday, April 3, one lousy payroll and dovish Federal Open Market Committee (FOMC) later, and all of a sudden the market was nearly certain the Fed will not hike in June. That was good for a 45 basis point rally in 10-year yields down to 1.80%. (2.25% was the gift).
It does not make sense for the 10-year Treasury first to sell off, and then to rally by 45 basis points, just because the market thinks the date of first fed hike has shifted by three or six months. Implicitly, there are big changes happening in the pace of tightening, the terminal rate, and in that catch-all fudge factor, the term premium, again, if we take as a given that the Fed will begin hiking sometime soon.
I think the market is finally coming to grips with the reality that a) it’s not a given that the Fed will hike “soon” and b) the Fed does not know what they will do. They appear to know what they want to do: plainly to get above zero on the policy rate. But now, living through a period of inadvertent tightening of financial conditions as a result of the rapid rise of the dollar, and commensurate weakening of the data, they are as anxious as the rest of us for the data quality to improve or persist long enough to justify policy normalization. They obviously scrutinize the data, and have human responses to it: “It’s giving me heartburn…” said James Bullard about last year’s 1st quarter GDP slump. Unfortunately, it’s probably happening again. The data weakness has caused estimates of first quarter 2015 GDP to be revised downward, perhaps to as low as 0.0% according to the undertakers at the Atlanta Fed1. Weakness was met with cheerleading in 2014, and we’re seeing it again now, with predictions that inflation will return to target and that the data will bounce back once transient factors (strong dollar, low oil, cold weather, port strike, etc.) dissipate.
In practice, it’s perfectly plausible to envision two scenarios. First, the recent poor employment report will be an aberration, and the strong trends will resume, bringing the labor market through the critical level of unemployment at which wage increases flare up in just a few months’ time; all while the negative transient factors dragging down the growth and inflation picture turn out to be just that, transient. I would argue it’s also perfectly plausible to envision a muddle through scenario, where the employment data remains fine, but that wage inflation is much further away due to structural changes in the labor market2 and that lackluster growth remains uninspiring. Would it really be so surprising to discover that above trend economic growth in the US is simply being restrained while the global hot potato of deflationary pressure gets passed around through the FX market, and now it is the United States’ turn in the hot seat?
This debate is now less about timing of the first hike and more about whether a meaningful tightening cycle is a given in the first place, regardless of whether the Fed can pull off a symbolic hike. Frequent readers will know my view, but the outcome is unknowable at this point, as a portfolio manager or as a policymaker wishing to retain the freedom to respond to new information. As long as this persists, so will the schizophrenic market reactions to each data print as we take a step closer to one outcome or the other.
2For those in the weeds, if the U6-U3 spread is a better ex-ante predictor of wage inflation than the non-accelerating inflation rate of unemployment (NAIRU).