Connect with us

Economics

JPMorgan Finds The Fed Has Broken The Most Fundamental Market Correlation

JPMorgan Finds The Fed Has Broken The Most Fundamental Market Correlation

A few months ago investors – especially those working for risk 60/40…

Share this article:

Published

on

This article was originally published by Zero Hedge
JPMorgan Finds The Fed Has Broken The Most Fundamental Market Correlation

A few months ago investors - especially those working for risk 60/40 balanced and parity funds - freaked out when the traditional correlation between stocks and yields (or inverse correlation between stocks and Treasury prices) flipped, sliding to the lowest on record, as any hope for diversification of equity risk by hiding into government bonds disappeared. And while the correlation has since recovered some of its normal historical pattern as the following chart from Goldman shows...

... a new and just as ominous decoupling has now emerged: that between stocks and investment grade and junk bonds.

In other words, while the low - or inverse - correlation between stocks and bonds has been one of the core anchors of modern finance, allowing cross-asset traders to diversify excess equity risk into bonds, this "basic premise" of modern portfolio consutrction theory no longer works. The culprit? Who else: the Federal Reserve.

As JPMorgan credit strategist Eric Beinstein writes in his latest Credit Market Outlook note, "in recent months, there have been two interesting trends in total return correlations: HG credit has become more correlated with stocks and thereby also more correlated with the HY bond market."

What's behind this rising correlation? According to the JPM strategist, there are several drivers:

  • First, the unusual dynamic that stocks are doing very well at the same time that UST yields have declined. This, in JPM's view, "is the result of so much QE-driven liquidity in the market that investors are buying everything: stocks and bonds."
  • The second, less likely, driver of persistently elevated correlations according to Beinstein, is exceptionally strong corporate earnings which have driven further total return gains this year alongside higher equities.

To JPM, this stands in contrast to what has traditionally been a key investment consideration for HG credit – its diversifying characteristic versus Equities, in the context of a balanced asset allocation portfolio (similar to the inverse correlation between stocks and treasurys). As Beinstein explains for those who have not taken finance 101, "when equities have risen, yields have often as well, leading to losses on the HG side and vice versa." But not anymore, and here's why:

The latest ongoing round of Fed QE appears to have broken this basic premise, with total return correlations between HG and Equities reaching their highest since 2008 on a 3yr trailing basis and since 1997 on a 12m trailing basis (of monthly returns in both cases).

The charts below shows these correlation time series over a very long period of time. They show that over time there have been various correlation regimes. Low to negative correlation has prevailed for most of the past two decades, but there have been several long periods of positive correlation in the past, with the most recent from 2009 to 2011, also following a market crisis.

Both charts above show that while there has been a wide range of return correlation regimes over the past 20 years "the increase in the Fed’s balance sheet (shaded area on the charts below) potentially argues for an extended period of higher correlations."

Some more observations on how this heightened correlation has played out in the past: HG returned -7.1% in March ’20, before the Fed Covid programs fully took effect, while equity markets weakened sharply as well. To JPM, with UST yields so low "there is a risk that this lack of correlation repeats, with spreads widening more than UST yields may be able to fall." Another takeaway is that the strong correlation between HG and HY returns argues for owning HY for the greater carry; HG and HY have been 75% correlated over the past 12 months, the highest since 2017. To be sure, it is always dangerous to extrapolate trends in correlation too far.

As Beinstein (somewhat sarcastically)concludes, "the potential implications of these developments are interesting" and explains: "traditionally portfolio theory says that bonds are a diversifier for equity market investments. This has not been the case recently, with the risk that it also remains not the case if/when there is an equity market selloff."

Translation: in the next crash, everything will go down at the same time and there will be nowhere left to hide... Which is also why the Fed can never again allow a market crash.

Tyler Durden Fri, 07/30/2021 - 12:30

Economics

Pros Increased ‘Crash’ Protection As Reflexive Vol-Sellers Rescued Stocks Yesterday

Pros Increased ‘Crash’ Protection As Reflexive Vol-Sellers Rescued Stocks Yesterday

A dramatic rebound in stocks – off the S&P’s 100DMA…

Share this article:

Pros Increased 'Crash' Protection As Reflexive Vol-Sellers Rescued Stocks Yesterday

A dramatic rebound in stocks - off the S&P's 100DMA - has prompted many commission-rakers and asset-gatherers today to call the end of the Evergrande event and signal the all-clear to new highs.

So what happened? What changed?

Nomura's Charlie McElligott explains that there is simply no way to overstate the power of the “reflexive vol sellers” into another spike, as this “sell the rip (in vol)” = “buy the dip (in stocks),” particularly as it related Put sellers either directionally shorting “rich” vols yday…and “long sellers” who monetized their downside hedges by the close (a lot of that being 1d SPY Puts from Retail “day traders” which doesn’t show in OI), creating $Delta to buy and again self-fulfilling yet another “turnaround Tuesday”

Critically, that Delta buying in the late day was hugely important then in reducing the absolute $ of systematic deleveraging “accelerant” flows, because only closing down -170bps in SPX then meant a much more manageable -$24.7B of Vol Control de-allocation in coming days, as opposed to what would have been a much more challenging -$62.9B to digest which we estimate would have been triggered off of a “-3% close”…while similarly, Leveraged ETFs only needed to rebalance -$5.9B at EOD, as opposed to a hypothetical -$8.9B assumed at the low of the day

Specifically, as SpotGamma details, the chart below shows that puts were net closed at all strikes above 4365 SPX (and 435 SPY) but there were fairly substantial positions added to lower strikes.

This indicates puts were rolled rather than outright closed. Again, with the Fed tomorrow trades want to leave some protection on.

Put volume surged relative to calls yesterday...

To Nomura's Charlie McElligott's amazement yesterday, we saw confirmation of our repeated point made stating that “the only things that clears out all that “crash” pricing in vol metrics is a crash”... yet it is VERY worth noting then that we actually saw Skew still steepen further yday despite incredibly high levels of both ATM Vol and Skew (SPX 1m 25delta Put Call Skew steepened 70bps, same gig for others: QQQ 64bps, IWM 37bps)...

...which tells us that the Dealer “short Vol / short Skew” problem still remains lurking in background.

SpotGamma concludes that its up to Powell tomorrow to set the next price move, which should be rather substantial due to the options positioning. Negative gamma could strongly influence any selling to the downside.

To the upside there is also a ton of fuel for an vanna-induced move if traders sell off their puts and crush the high implied volatility levels. Therefore while today is likely about chop, the move out of Wednesday should be substantial.

Tyler Durden Tue, 09/21/2021 - 09:49
Continue Reading

Economics

TSX down in every sector amid Canada election & Evergrande crisis

The substantial bankruptcy by Chinese property developer Evergrande triggered an economic slowdown on Monday September 20 across the globe which is…

Share this article:

The substantial bankruptcy by Chinese property developer Evergrande triggered an economic slowdown on Monday, September 20, across the globe, which is impacting the financial systems. Additionally, the upcoming Fed’s action for pumping stimulus into the economy seemingly created uncertainty among the investors.

All the 11 major sectors registered on the TSX composite were down including healthcare (5.01%), Energy (2.84%), Industrial (1.78%), Financial (1.75%) and Technology (1.51%). Thus, the TSX composite index continued in the red zone after losing 335.82 points or down by 1.63% settled at 20,154.54.

The one-year price chart (as on September 20, 2021). 

Active volume

TC Energy Corporation was the most actively traded stock where 17.80 million exchanged hands, followed by Canadian Natural Resources where 15.02 million exchanged hands, and Suncor Energy Inc. with 8.64 million shares exchanging hands.

Movers and laggards

Wall Street update

Concerns over the probable collapse of China's real estate company Evergrande weighed heavily on US equities on Monday, with the main averages all suffering their biggest losses in many weeks.

Traders were also looking forward to the Federal Reserve's monetary policy statement on Wednesday. The central bank is largely anticipated to maintain its current monetary policy, but it may discuss the future of its asset purchase program.

The Dow Jones Industrial Average finished the afternoon down 614.41 points or 1.78% at 33,970.47. The S&P 500 fell by 75.26 points or 1.7% to 4,357.73, while the Nasdaq was down by 330.06 points or 2.19% to 14,713.80.

Commodity update

Gold traded at US$ 1,763.80, up 0.54%.

Brent oil slipped down at US$ 73.92/bbl down 1.48%, while Crude oil also traded at US$ 70.29/bbl down by 1.72%.

Currency news

The Canadian Dollar slid against the U.S. Dollar on Monday, while USD/CAD closed at 1.2825, climbing 0.46%.

The U.S. Dollar index gained marginally against the basket of major currencies on September 20, and ended at 93.23, up 0.01%.

Continue Reading

Economics

Transitory-Ness In The TIPS Market

With all the debate about the persistence of inflation, one natural thing to ask is: what is the market pricing? Unfortunately for those who are selling a “return to the 1970s” narrative, “Mr. Market” is firmly in the camp of “Team Transitory.” Of cour…

Share this article:


With all the debate about the persistence of inflation, one natural thing to ask is: what is the market pricing? Unfortunately for those who are selling a “return to the 1970s” narrative, “Mr. Market” is firmly in the camp of “Team Transitory.” Of course, one imagines that the immediate response is that the “markets are wrong.” Since many of the people in the “inflation is coming!” crowd are also in the “markets are always right” camp, there is a good chance they might modify the argument to “markets are wrong according to some affine term structure model.”

(Note: The server issues that stopped me from posting this yesterday are obviously sorted.)

The baseline for our discussion is the most straightforward measure of “forward inflation expectations” is the figure above: the 5-year TIPS inflation breakeven rate, starting 5 years forward. I took the series directly as calculated by the St. Louis Federal Reserve as based on the Fed H.15 fitted Treasury rates, but they used the exact same approximation I would use in this circumstance.

To give a quick primer for those new to the topic, the “breakeven inflation rate” on an inflation-indexed bond (in this case, U.S. TIPS) is that future rate of inflation that gives the same total return as nominal government bonds (conventional Treasuries). That is, it is literally the rate of inflation where the two types of government bonds have the same return (break even). From the perspective of mathematical finance — as used in option pricing — this is where the expected inflation rate is, using the mathematical definition of expectation. One should keep this notion of “expectation” distinct from the notion of “forecast,” as I discuss below.

I showed the forward rate instead of the spot rate (a breakeven over the time period starting now) so that we can see what is priced in after current disturbances have subsided. (I did not even bother looking at spot breakeven inflation rates.)

As the chart above shows, the forward breakeven has recovered from the usual collapse we see in financial panics during the lockdown period, but are still well below where it was last cycle. (Which was too high in retrospect.).

(My book Breakeven Inflation Analysis offers more information on this topic.)

I would argue that the only sensible takeaway from that figure is that inflation market participants are not deeply concerned about a secular rise in inflation. Beyond that, I do not have enough information to be more specific. My immediate concern is that the above time series is an approximation, and we really would need to dig in the details of bond-specific pricing to have more confidence in the exact level of forwards.

The reality is that TIPS market is not incredibly liquid, and there are fewer issues to work with. Meanwhile, an individual issue can have pricing quirks like seasonal effects based on its coupon dates. The figure above certainly gives the correct big picture, but one should not view the specific levels as being definitive at any point in time. Even without access to the instrument-specific data, I believe that it is fair to assume that error bars are currently larger than usual.

(The inflation swap market can give a cleaner read on forwards, but I am out of the loop with regards to the liquidity in that market currently. Inflation swaps always tend to be secondary to government bond inflation markets, since it is extremely hard to find anyone other than the central government willing to sell inflation protection in size.)

The reason why academics or central bankers will say the above reading is incorrect is because they have models that allege that the risk premia have been zooming around, and so the “real expectations” are way higher than observed breakevens. That is, the breakeven might be the mathematical expectation, but forecasts (“expectations” in common English) are higher.

One of the interesting things about fixed income is that outsiders insist that bond market participants routinely misprice bonds while within the bond market there are entire teams whose entire day job is to price bonds correctly. (I used to lead one of such teams.) Whatever.

The problem with the “risk premia zooming around” story is that it is unclear that any of the models offer useful information. They are internally consistent, yes. But do they offer useful information — where we measure “usefulness” by generating profitable trade analysis? In general, they do not pass that test. For example, it is entirely typical for the models to suggest that term premia in a certain part of the curve was heavily negative — and then the realised excess returns were quite positive, as any ignorant shnook fixed income quant could have told the academics.

In this case, I have some sympathies for a demand for inflation protection forcing breakevens lower. But there is no way to get from there to a “1970s is coming!” narrative. At best, we are “really” just back where expectations were last cycle.


Email subscription: Go to https://bondeconomics.substack.com/ 

(c) Brian Romanchuk 2021
Continue Reading

Trending