This time is different. No, really!

The phrase “this time is different” doesn’t usually spark very positive reactions. But I don’t care because there’s one thing about the market these days that makes me think that something is strange when compared to previous bond market sell-offs.

If you follow me on twitter you will have noticed that lately I have been talking a lot about asset swaps (ASW). This is a pretty technical concept but I will try to be as straightforward as possible.

If you are still reading this then you probably know that in the fixed income market we have two broad groups of instruments – cash bonds and swaps. In theory, their yields (prices) should be moving more or less in a parallel fashion because they are interest rates instruments. In other words, you can bet on interest rates going lower by either buying bonds or receiving interest rate swaps (IRS). The difference between those two instruments is called the asset swap and it tends to move for the following main reasons:

  • Buying bonds requires balance sheet and IRS is an off-balance sheet instrument.
  • Buying government bonds creates exposure against the issuer (sovereign) while IRS is a contract between two counterparties (e.g. banks).
  • Supply of bonds is limited while IRS can be created out of thin air.
  • Government bonds are a stream of cash flows (coupons) while IRS is an exchange of fixed against floating rate (e.g. LIBOR).

I have learnt to pay a close attention to moves in ASWs, just like I very closely monitor moves in cross currency basis because they can reveal pretty significant market developments. An ASW can tighten, i.e. bond outperforms the swap (e.g. bond yields drops by 10bp and IRS for the same maturity by 7bp) or widen. In core markets, tightening of ASW has been historically connected with higher aversion to risk. When problems arise, investors would very much rather own, say, US Treasuries than have a contract with a bank to exchange some cash flows. Chart below shows the 10y ASW (the higher the number, the more expensive the bond vs swap) in the US against the EMBIG spread. As you can see, the correlation is pretty significant.


Now, what happens in developed markets does not usually work the same way in emerging markets. Indeed, periods of risk aversion were generally associated with significant widening of ASW in emerging markets. The rationale is simple – let’s dump emerging bonds because the credit risk is going up. Having an interest exposure via a swap with JP Morgan becomes more valuable than buying government bonds of governments of Mexico, Hungary or Malaysia. Simple heuristic.

And this brings me to the “this time is different” proposition. As you may have noticed we are experiencing the end of days for government bond markets. Well, we’re not really but people like Bill Gross want to make you think like that. EPFR data is showing significant outflows from bond funds investing in emerging government bond markets. The last time we saw such big outflows was in September 2011. However, unlike in September 2011 when ASW totally exploded, in the recent weeks EM ASWs have actually tightened and considerably so. Just to give you an example – ASW in the 10y segment of the South African government bonds are at the tightest level they’ve ever been. South Africa – the country whose economy is in a downfall, whose currency has sold off dramatically and where the social tensions are at levels unseen in years. To be sure, the bonds have sold off too but nowhere near as much as IRS.

I can find a few explanations for that but the most important conclusion is that real money investors (so asset managers rather than hedge funds) have not been selling government bond markets to a large extent. They have sold some and shifted others to more defensive places, they probably hedged their currency exposures but they have not sold their bonds. Why? Perhaps because they don’t believe Bill Gross, thinking the scare will pass (this argument seems to be supported by Pawel Morski in one of his latest posts). Or perhaps because they know the market is not able to absorb the potential flow anyway.

At the same time, the hedge funds seem to be willing to exploit the recent change in the global mood and are pushing IRS higher. This then stops out model accounts (CTAs, aka the scum of the earth), which had been running humongous receiver positions in bellies of various curves assuming the Fed would stay put forever (or at least 3-5 years). Meanwhile, it seems like the tide has turned a bit and the convexity of the US curve is shifting. That’s why I was pointing out earlier today this tweet from Business Insider’s Matt Boesler:

All in all, I don’t want to make this post too technical but this is the first EM bond “crisis” since I have been in the industry when local bonds in emerging markets have so far been outperforming IRS and ASW have been tightening. And while I think I understand the reasons behind that this is not a sustainable situation, in my opinion. In fact, I strongly believe that something has to give – either the real money guys are in a denial or the hedge funds have jumped on the tapering bandwagon too early. Either way, the EM curves are pricing something that is almost impossible to come true, in my view.

Inflation volatility or which bond markets to avoid

As promised, a quick word about inflation linkers in emerging markets.

While in most cases, the path of inflation in the coming 3-6 months is pretty much given; many economies are a completely different level of their respective inflation rates. Importantly, volatility of inflation rates also differs wildly. Take the following chart as and example. It shows the relationship between average inflation since the beginning of 2005 (so around 100 observations) and the standard deviation of CPI in the same time-frame.


As you can see, there is a group of emerging economies which have achieved a relatively low level of inflation (the circled part of the chart) but they have had mixed success in terms of staying close to the average. Arguably, Israel, Malaysia and Poland are the best in this field.

As you move to countries with higher average inflation rates, standard deviation tends to rise. However, note that we are talking here in percentage points, as opposed to normalised values. In other words, we say that Poland is “better” than Turkey just because it has had a lower average and a lower standard deviation. However, if you normalised standard deviation by dividing it by the average you will see that Turkey has been much closer to its average than Poland. We will use that in a minute.

The chart below combines the approach of the volatility in inflation rates (i.e. standard deviation divided by the average) with the distance of the latest CPI print from respective central banks’ targets (note that for some countries I had to make assumptions as their central banks are not inflation targeters).


In this chart we see, for example, that South Africa is 1.2pp above the 4.5% target (mid-point of the 3-6% SARB range) and  that standard deviation of CPI prints since 2005 has been around 40% of the average (which was 6.1%).

How to trade this? Someone could ask what the point of looking at such a chart is. Well, as worries about inflation resurface along with some acceleration in growth rates, investors will be willing to bet that some emerging economies could have their inflation rates moving up fast. At the same time, the central banks would have to respond. Therefore, I think that what we’re looking for is countries, which

  1. currently have inflation below (or close to) the central bank’s target
  2. have experienced significant volatility of inflation prints in the past.

In those economies you should consider looking at inflation linkers or shorting nominal bonds.

When using such a comparison, Chile stands out as a good candidate with low current inflation and high inflation volatility. Similarly, Israel has a history of quite rapid moves in inflation rates and we also can be reasonably sure that the output gap there is insignificant. Finally, Poland is looking at rapid declines in inflation rates at the moment, mathematically increasing odds of a rebound in the second half of the year and indeed in 2014.

There are also many caveats to this approach but I have found it to be a useful starting point when trying to play a global inflation / disinflation theme.