Alpha Exchange is a podcast series by Dean Curnutt to explore topics in financial markets, risk management and capital allocation in the alternatives industry
- Tweets 5,673
- Following 1,106
- Followers 21,999
- Likes 103
Pinned Tweet
Charts referenced in most recent pod:
“The Market Disregards Correlation”
Apple:
podcasts.apple.com/us/podcas…
Spotify:
open.spotify.com/episode/772…
2
2
38
8,865
big believer in financial innovation and the freedom to create and choose. at the same time, frustrated with the current state, which is more alchemy than innovation. as the kids would say, these products coming to market are "NTB".
2x leveraged ETFs on crypto treasury companies, autocallable ETFs with fantastic embedded margin for the seller... products that simply do not serve the end user. The latest is that Robinhood is creating 15 minutes(!!!!) contracts on Bitcoin.
1
10
1,638
Good am. It certainly doesn't feel like a realized volatility environment where 2 week is 12 and 1m is 11 on the SPX. It's difficult not to stare at 10's and 30's all day long.
Lots of takes on the “why” of the long-end sell-off. Attribution is a most popular sport on Wall Street. Let’s see….in no order and probably not complete…
1. Fiscal irresponsibility premium
2. (Related) never-ending supply
3. Inflation above target for 65 months
4. Short rates moving higher in response to onset of tightening cycle
5. #2 but inclusive of AI related issuance
6. #3 focused on the war and crude
7. Strength of economic growth
8. The trend is simply for higher yields globally.
On this last point, the average yield on government debt worldwide sits just shy of 4%, the highest since 2007
That’s a lot of reasons, none of which should be ruled out. In combination, it’s easy to justify the move, with the potential for more to come.
There are many takes as well on whether we are on the precipice of a crisis. We can at least agree that the circumstance is fragile, I hope.
The correlation between stock and bond returns is at or near a record high. The back-end is the main threat to the equity market.
Resharing this chart from yesterday which I really think nails the two incredible outliers in correlation.
Stock to Bond: 100th percentile
Stock to Stock: 0th percentile
The correlation among risky assets is considerably LOWER than the correlation between risky and risk-free assets. The case for options-based insurance is easy to make, it's just about stomaching the challenging carry.
3
3
24
2,225
If there's one chart that frames the challenges to risk management right now, I think it's this one.
Going back 15 years, here's the rolling 6m realized correlation between the $SPX and $TLT as well as the correlation among stocks in the SPX.
In the "risk on/risk off" era, stocks and bonds were vastly negatively correlated. No longer. The latest reading is a positive 46%.
In that same era, stocks were consistently and meaningfully correlated to each other, reaching as high as 80% in crisis periods like the GFC, 2011 Sovereign Crisis and the Covid unwind. That's a thing of the past as well. The latest reading is 5%.
Stock to Bond: 100th percentile
Stock to Stock: 0th percentile
The correlation among risky assets is considerably lower than the correlation between risky and risk-free assets.
Many nominally different assets - utilities and tech stocks, for example - are not correlated today, but are linked to a common factor that could drive correlation in the future. As the bond market is as much a threat to the stock market as it is a flight to safety asset, the importance of finding real diversifying assets is critical.
3
8
42
6,739
“Financing the AI Buildout”.
Latest podcast features Amanda Lynam, Chief Credit Strategist at Goldman Sachs.
We discuss the insatiable demand for debt capital from hyperscalers and AI-centric corporates and where the funding will come from. Amanda also shares her views on sources of diversification for credit investors, using exposure to sectors like healthcare.
axpod.com/podcast/amanda-lyn…
12
2,242
I dopped this "solo pod" a few weeks back, making the case that the cost of hedging vulnerability is low to the degree of vulnerability. Bond market is central to the threats to the equity market.
axpod.com/podcast/the-case-f…
1
18
2,428
the old adage, "the cure for higher prices is higher prices" really applies to rates. rates move up enough and they plant the seeds of a risk-off significant enough to lead to lower rates.
3
4
58
20,323
Here's a table with top 10 increases in the 10y yield since 2016 and the MOVE on the way in and out. We might eliminate the two that occurred in Mar'20 for special circumstance.
The 2022 tightening cycle has a bunch of hits. Note that the MOVE was below 100 only twice.
2
1
19
2,276
Table below shows buckets of observations of the $VIX and $VVIX (the VIX for the VIX). These measures are certainly correlated to each other. A spike in vol also typically leads to a spike in vol of vol.
Conversely, when vol is quite low, it's generally the case that the VVIX is low as well. An exception would be 2017, when the VIX averaged 11 and the VVIX averaged 90, a ratio of 8x.
Today, the VIX sits in the 4th percentile of observations over the past two years. The VVIX sits in the 1st. In combination, these inputs make for excellent economics on VIX call options.
No doubt, these trades are being put up at a time when the feedback from carry has been especially challenging. Realized vol on down days is 6.8% over the past month. That's crazy low. We are 103 days since a down 2% day (June 5). Markets at rest do tend to stay at rest, so we do know from experience that low vol periods can persist.
2
1
26
2,248
Vol at the index level is a joint function of the vol of the stocks in the index and the correlation among them. It's the incredibly low level of the latter that has been Ozempic for index volatility.
"Low correlation each day keeps index vol at bay"
Here's a chart that shows it. Assume 35 for the vol of the average stock in the $SPX. With the $VIXEQ at 38, that's a reasonable assumption. Now choose different correlation levels (horizontal) to yield different index vol outcomes (vertical).
Actual 3m realized correlation of the stocks in the SPX is (wait for it) 3%. One year is 7.5%. Let's move correl from 7.5% up to 30%. SPX vol essentially doubles, up from roughly 10 to 20.
And that assumes no change in average single stock vol. We know, empirically, that stocks become more volatile and more correlated at the same time. A joint shock higher to both will seriously boost realized vol at the index level and take the $VIX much higher in the process.
13
10
101
26,900
some part of the "stickiness" of the $VIX despite a good pop in the $SPX today is a function of the shape of the S&P 500 volatility surface.
Remember, the VIX is an engine that grabs new strikes to do its calc each time the SPX moves. when the new strikes have a much different vol than the old strikes, you get a lot of VIX movement.
in today's circumstance, the "call skew" on the SPX is very flat. that means that the implied vol of the new strikes that become part of the VIX calc after the SPX has risen are not too much different than the prior strikes.
in the chart below, I show the top 10 VIX *declines* on days the SPX has risen between 1.25 and 1.75%. two of those dates are graphed (3/6/07 and 6/23/16). note the big difference in the shape of the vol surface of those two versus today. the rally in the market on those days moved us to considerably lower vols for the new strikes. today's vol surface is much flatter so the VIX moves less.
2
1
27
3,039
As the next Fed meeting (10/28) occurs and right around the mid-term outcomes (11/3), we'll see a raft of mega-cap earnings. In just two days, we get META, GOOG, MSFT on 10/29 and AMZN and AAPL on 10/30. That's 25% of the market cap of the SPX.
What you can see below is how the market prices the "event vol" from earnings via the hump in implied vol between 1m, 2m and 3m.
1
1
14
2,307
Wrote this late last year. Clearly the “correlated unwind” hasn’t occurred but do see more discussion of the feedback loop between the market and economy.
The market imposes risk on the economy.
TLDR: the economy --> market feedback loop today is similar to that of the pre-GFC period. It's the market that will take the economy down, not vice versa as is traditionally the case.
Today's SPX can be summarized as "highly concentrated with highly volatile, highly valued but remarkably uncorrelated tech stocks".
A good argument can be made that the market is not properly identifying the linkages, cross-holdings, investments, and extent to which customer/supplier relationships underpin the correlation in outcomes for AI focused stocks. We are still in the leveraging period and stock price changes have been vastly idiosyncratic.
A similar argument could be made for home prices in pre-GFC era. Housing price appreciation was clearly driven by a common factor: the bottomless extension of mortgage credit. But that did not show up in city to city correlations until there was a break in the circularity. Once defaults picked up, the credit machinery failed and the correlation of housing prices surged.
In the aftermath of large drawdowns, investors consistently realize they’d underestimated the degree of “sameness” in assets. It took us until 2008 to recognize that the huge run up in housing prices was linked to a common driver: the vast supply of mortgage credit.
Today, are we missing the vulnerability to a Mag7 sell-off? The negative wealth effect would be substantial. If market cap is the “currency” to fund Capex and that same Capex is driving economic growth, a sell-off in Mag7 has multiple pathways for spill-over.
Below, a few pieces picking up on the same idea.
1
2
20
4,225
I periodically repost this chart, a favorite. It’s the pathway to zero from being long vol (UVXY) and short vol (SVXY)
“When do you wanna go bankrupt, now or later?”
7
10
136
15,763
We've got a great fall schedule for the podcast. Lucky I can put myself amidst these accomplished professionals!
axpod.com/wp-content/uploads…
1
8
3,002
the single best quote and lesson coming out of the SALP unwind is from @KrisAbdelmessih ... "never give a 24 year old money"
here's the potential footprint today...punting on 50k of far upside calls on NBIS in the FLEX market, paying 90 vol. bought 2.4mln of vega, paid more than 110mln in premium.
15
21
576
324,226
Thank you @fejau_inc and @ForwardGuidance for having me on your podcast.
Option nerds like me live in the tails of the distribution, asking "what's unpriced?" A correlation event in equities is almost entirely disregarded. That's not to predict that it will occur, but that you can buy the potential that it may at a very cheap cost.
nitter.net/ForwardGuidance/status…
NEW POD
We Cover:
🔸 Why stock correlations are so low
🔸 If dispersion trades unwind
🔸 Systematic tail hedging
🔸 AI concentration amplifying risk
🔸 If the Fed can calm markets & more!
@fejau_inc @Dcurnutt @Alpha_Ex_LLC
TIMESTAMPS:
00:00 Intro
05:08 Why Stock Correlation Collapsed
10:49 What’s Driving The Dispersion Trade?
18:45 Ads (Token 2049, Avalanche Summit)
20:21 Could Volmageddon Happen Again?
27:16 Why Tail Hedging Looks Attractive
35:35 Can Portfolio Insurance Ever Be Free?
39:55 Systematic Or Discretionary Hedging?
42:33 Where Could Market Risk Emerge?
47:27 Can The Fed Calm Markets?
53:22 Closing Thoughts
26
4,692
$VIX at least 35….so many asset classes and sources of risk…
1997 - SE Asia FX implosion
1998 - Russia default, swaps, vega
2001 - terrorist attack
2002 - accounting fraud, credit events
2008 - subprime CDS, banks
2009 - subprime CDS, banks
2010 - Flash Crash
2011 - US debt ceiling
2015 - China FX reprice
2018 - VIX ETP meltdown
2020 - global pandemic
2022 - rate shock
2024 - Yen unwind
2025 - tariffs
2
2
34
3,792