How Ugly Will It Get: CTAs Are The Most Short Treasurys Since 2018… And Getting Shorter
A little over a month ago, on January 7 when looking at the technicals in the Treasury market and when the 10Y was trading at just over 1%, we warned that “it’s about to get ugly”…
10Y Treasury Yields Hit CTA Liquidation Trigger: It’s About To Get Ugly https://t.co/0Zll3ERxGr
— zerohedge (@zerohedge) January 7, 2021
… revealing that the key catalyst for further downside in rates would be when trend-following funds such as CTAs liquidated their long positions and flipped from long to short, an event which Nomura’s Masanari Takada said would take place when “10yr UST yields remained above 1.10%”…. which they clearly have in the past few weeks.
Well ,since then it has indeed gotten quite “ugly”, with the 10Y blasting off from 1% to an intraday high above 1.43% today (at which point Powell was forced to talk down rates once again this morning).
And since the sharp move higher in yields has already had a sharp and cascading effect on risk assets in general and high-duration tech and growth stocks in particular, the question everyone wants answered is what happens next, and will CTAs keep hammering the 10Year?
To answer that question we first need to know how positioned CTAs are currently – after all, one month ago we warned that their liquidation and subsequent short reversal would be the catalyst for the next big move.
Well, as JPM’s Nick Panigirtzoglou writes overnight when he confirms just what we said back in January, “the bond market sell-off has likely been amplified by CTAs” and substantially at that: according to JPM estimates, CTA and momentum signals have shifted to their most bearish territory since 2018, although they are still some way from extreme territory (which is probably not good news for those hoping for the shorting to be over).
According to the JPM strategist, “the last week or so has seen an important shift in our signals. Figure 6 depicts the average of the shorter- and longer-term signals for 10y Bunds and USTs, and shows that both have now turned bearish. Indeed, the signal for 10y Bunds turned short on Feb 12th and is now sitting at its most bearish level in two years. Similarly, the signal for 10y USTs turned bearish on Feb 16th and is now sitting at its most bearish level since late 2018.”
This confirms what we warned about in January, and to JPM, “suggests that CTAs have served to amplify the bond market sell-off in recent weeks.”
Unfortunately, the fact that CTAs have aggressively increased their bearish bias yet are well off extreme positions, cuts both ways: while they could in theory reverse, they can just as easily press momentum and push the 10Y to the critical 1.50% which according to Nomura – which has been well ahead of everyone in its analysis of bond-risk dynamics – will mark the level where stocks suffer a substantial selloff.
So what does JPMorgan, which is once again well behind the curve, think? As Panigirtzoglou writes quite belatedly, the bond market sell-off over the past weeks has been accompanied by a drift up in bond-equity correlation, “raising concerns about de-risking by multi-asset investors, such as risk parity funds and balanced mutual funds” something we also briefly mentioned on monday when the big Treasury puke first emerged:
Not a good day for risk parity
— zerohedge (@zerohedge) February 22, 2021
Since these two types of investors benefit from the structurally negative correlation between bonds and equities (as this negative correlation suppresses the volatility of bond-equity portfolios allowing investors like risk parity funds to apply higher leverage and thus boost their returns and improving risk-adjusted returns for balanced mutual funds as equity drawdowns are mitigated by bond allocations) the opposite takes place when this correlation turns positive: the volatility of bond-equity portfolios increases, inducing these investors to de-risk. This overall increase in risk portfolios has been apparent in recent days with both equity VIX and the bond volatility index, MOVE, rising sharply.
So “How worried should we be about de-risking by these two types of investors” JPM asks?
As shown in the chart below, there were a few episodes over the past years during which the bond-equity correlation turned positive. The major ones were five: the Fed taper tantrum of May-June 2013, the Bund tantrum of May-June 2015, the period into the US election Oct-Nov 2016, Feb 2018 and Q4 2018. Outside these episodes the bond-equity correlation has been predominantly negative helping to contain the volatility of risk parity funds and balanced mutual funds.
What have we learned from those episodes, JPM asks rhetorically? Which ones were most problematic for risk parity funds and balanced mutual funds? Figure 2 and Figure 3 show the performance of these two types of investors during the above episodes when the bond-equity correlation turned positive. For the $150bn risk parity fund universe, the most problematic episode was the taper tantrum of May-June 2013 as they had entered that episode with very high leverage. As a result they were forced to de-risk abruptly and suffered a heavy 10% loss. The other episodes, i.e. May-June 2015, Oct-Nov 2016, Feb 2018 and Q4 2018 were relatively less severe, as risk parity funds lost around 4% and their de-risking was likely more orderly.
What about balanced mutual funds, a $1.5tr universe in the US and $7tr globally?
According to the Greek strategist, these funds have typically higher allocation to equities, around 60%, and are less mechanical and thus have more flexibility in responding to changes in volatilities and correlations. As a result of this flexibility and their higher equity allocation, they posted modest losses during the episodes of May-June 2013, May-June 2015 and Oct-Nov 2016. However, they suffered heavy losses during the Feb 2018 and Q4 2018 episodes as they were caught up with higher equity positioning and thus forced to de-risk by more than the former episodes. This shows how important is the starting point of risk positioning in determining the eventual loss in each episode.
Finally, how vulnerable are these types of investors currently in a scenario where the bond-equity correlation continues to creep up taking the form of a tantrum similar to the above episodes, whether with the involvement of CTAs or not?
To assess their vulnerability JPM resorted to two position proxies. For balanced mutual funds it compared their returns to that of 60:40 equity:bond benchmark as shown in Figure 4. For risk parity funds the bank gauged leverage by using the ratio of their volatility to the volatility of a risk parity fund benchmark. This risk parity fund leverage proxy is shown in Figure 5.
Figure 4 and Figure 5 show that balanced mutual funds look more vulnerable than risk parity funds currently. This is because after de-risking in January these balanced mutual funds appear to have raised their risk positioning, i.e. increased their equity overweights and bond underweights, to high levels in February. Instead, after de-risking in January, risk parity funds’ leverage stayed below average during February. In other words, if the bond-equity correlation continues to creep up, e.g. via a worsening of the equity market drawdown even as yields continue to grind higher, balanced mutual funds pose a greater vulnerability for the equity market.
Finally, JPMorgan cautions that the elevated positioning by balanced mutual funds “is also raising questions about month-end rebalancing” as the equity rally and the bond selloff during February “is creating a pending rebalancing flow for balanced mutual funds (assuming the past few days equity market correction does not propagate to erase any pending rebalancing).”
Assuming balanced mutual funds had fully rebalanced by January, which is a reasonable hypothesis given the reduction in their betas in January, JPM expects around $90bn of equity selling by them for February month-end. However, as we discussed last Nov/Dec and as some – namely those who have a habit of trying to frontrun month and quarter-end rebalancing – learned the hard way is that balanced mutual funds do not necessarily rebalance every month. In fact, according to JPM, during the previous quarter, they appear to have postponed rebalancing during Nov-end or Dec-end and to have waited until January to de-risk/rebalance. As such, in JPMorgan’s opinion given balanced mutual funds had de-risked in January, it would be too soon to rebalance again in February, and the bank believes that “they will likely postpone any pending rebalancing to March.” On the other hand, considering that this advice comes from JPM which has been painfully wrong about most market trends and themes in recent months, don’t be surprised to see a violent and sudden month-end flush…
Wed, 02/24/2021 – 13:37