Stocks are learning to “look through” and otherwise “write off” the worst data in the history of modern economic statistics.
By now, we’ve all come to terms with the self-evident notion the data aimed at capturing and quantifying economic activity cannot possibly be any semblance of robust when economic activity is forbidden by government decree.
Hollywood
Those of an empathetic persuasion remind themselves that when it comes to things like jobless claims, numbers are actually people. But if you remember 2008, and you’ve got capital to deploy or reallocate, you’re worried you may miss the opportunity of a lifetime. How foolish will you look (or at least feel) in retrospect, knowing you sat idly in cash despite the single largest monetary/fiscal “bailout” in history staring you right in the face?
And what will be your rationale? You’ll have to explain, if only to yourself, that your penchant for caution was based on the fear that a pandemic may ultimately metamorphose into a Hollywood-style zombie blockbuster.
Maddeningly for market participants, the Hollywood doomsday outcome isn’t as totally far-fetched as it usually is. Ironically in that context, LA went from 1,091 film shoots in February to zero last month, Bloomberg reported on Thursday afternoon.
I remember where I was when the world didn’t end a dozen years ago. And now here we are again pondering the apocalypse. Only this time, there’s real pestilence.
After easily brushing aside a series of laughably bad PMIs and another week of massive jobless claims which, with a little extrapolation, suggest that less than half of working age Americans will be earning a wage next month, Wall Street stumbled over ostensibly bad news around Gilead’s remdesivir on Thursday. The drug has been billed as a potentially promising therapeutic in the fight against COVID-19.
Trial by fire
A summary of a Chinese trial suggested the drug was a “flop” (as the media put it), a characterization Frederick Hayden, an infectious disease expert at the University of Virginia School of Medicine who assisted Chinese doctors with the study, called “not correct”. “My interpretation… is not consistent with that headline”, he said.
But what does he know, right? He’s just a scientist involved in the actual trial. The real experts are over at the Financial Times (which originally ran with the “flop” headline).
Baird analyst Brian Skorney piled on, calling this the “worst case scenario as not only did the clinical effect not manifest, but there wasn’t even an antiviral effect to explore”. The results, he said, are “very sobering”. The shares slumped 4%.
Making matters worse, the abstract of the study was mistakenly posted by the World Health Organization, which is now destined to be blamed for another blunder associated with the virus just a week (give or take) on from having a serious chunk of its funding cut off by Donald Trump.
Gilead called the summary misleading. “Trends in the data suggest a potential benefit for remdesivir, particularly among patients treated early”, the company said, in a statement, adding that it’s potentially erroneous to suggest that conclusive results can be drawn from studies with low enrollment.
Both Gilead and Hayden said the results of the study are (or will be) under review for publication.
So, you’re a rocket scientist?
Actual COVID-19 cases in the US rose the least in a month on Thursday. At the same time, nearly 14% of 3,000 New Yorkers tested in a state program were positive for antibodies to the coronavirus, Andrew Cuomo revealed.
More than a fifth of those tested in New York City (21.2%) had the marker.
The implications are clear. Some 2.7 million New Yorkers may have been infected. That is 10X the official, confirmed total, as of Thursday.
This means the fatality rate is almost surely dramatically lower than reported – around 0.5%, Cuomo said, although that may not capture scores of unfortunate souls who have died at their homes without being diagnosed.
“Howard Zucker, the state health commissioner, said New York’s test had been developed in a state lab that had been given blanket approval for such tests by the F.D.A.”, The New York Times writes, adding that although Zucker acknowledged New York City’s concerns “about some tests on the market, the state’s test is reliable enough to determine immunity – and, possibly, send people back to the office”.
Dr. Demetre Daskalakis, New York City’s top disease control official, warned that the tests and others like it can’t be relied on to project “durable immunity”. “Serologic tests should not be used to diagnose acute or prior SARS-CoV-2 infection, nor should they be used to determine immune status to SARS-CoV-2”, Daskalakis wrote, in an advisory. “They may produce false negative or false positive results, the consequences of which include providing patients incorrect guidance on preventive interventions like physical distancing or protective equipment”.
He was quite explicit. “Given the current lack of evidence that detection of SARS-CoV-2 antibody on any serologic test is indicative of durable immunity, it should not be used for that purpose”, he said, in the same linked health alert.
As The Times goes on to write, “the guidance from the city’s Department of Health appeared to open another rift between the city and state government”.
“It is a way to say this person had the disease and they can go back into the work force”, the above-mentioned Zucker remarked, defending the tests. “A strong test like we have can tell you that you have antibodies”.
In addition to showing that more than 1-in-5 New York City residents tested positive, the state’s survey indicated the rates were 16.7% for Long Island and 11.7% for Westchester and Rockland counties.
Consider those figures (21.2% for New York City and the county-level numbers) and compare them with those shown in the following table.
(JPMorgan)
That figure is from a note dated April 13 by JPMorgan’s Marko Kolanovic. I discussed the analysis at length the day it was released in a post called “Marko Kolanovic Addresses ‘Critical’ COVID-19 Questions On Herd Immunity, True Mortality Rate“.
The figures unveiled by Cuomo on Thursday match Kolanovic’s forecasts for New York City, Rockland, and Westchester almost exactly.
The JPMorgan quant (a household name, as far as sell-side analysts go) derived his projections by backing out infection and mortality rates using examples from around the globe. Again, you can peruse the details in the linked post, which includes discussions of outbreaks in Gangelt, Germany, and the Faroe Islands, two interesting and highly informative case studies.
Prior to the publication of the forecasts shown in the table, Kolanovic published a series of notes modeling a variety of crucial COVID-19 outcomes, including, but not limited to, hospitalization rates, herd immunity and mortality rates. He also sketched out the broad contours of a national plan for pandemics, which would entail harnessing big data and machine learning in order to help humanity stay ahead of an outbreak.
In a separate note dated April 6 (discussed here) Kolanovic said New York had likely seen “peak cases” on April 2 and would begin the recovery phase on April 19. Those projections were based on his own models as well.
Generally speaking, that has proven to be more or less correct. In fact, on April 19, Andrew Cuomo said the following:
This is the state of hospitalizations. We’ve been watching this 24-hours a day for it seems like most of our lives, but it’s only been about 40 days. The total hospitalization rate is down again in the State of New York. We’re down to 16,000. If you look at the numbers, we were at 18,000 people hospitalized for a period of time. It flattened there for a while, it paused there. Then it went down to 17,000 but this is a low from our high point of 18,000.
Big question of whether we’ve been past the apex, past the high point and it turned out the high point wasn’t a point. The high point was a plateau and we got up to a high point and then we just stayed at that level for a while. If the data holds and if this trend holds, we are past the high point and all indications at this point that we are in a descent.
Take a moment to let that resonate. On the very same day that Kolanovic predicted New York would enter what he called a “recovery phase”, Andrew Cuomo (cautiously) said the state is “past the high point” and “in a descent”.
Additionally, Kolanovic predicted in March that hospitalizations would peak at a fraction of then frequently-cited models. “Our model shows peak hospitalization occurring this past weekend”, he wrote, in a note dated March 30, adding that “we don’t rely on the total number of hospitalizations in our model as accurate, as it needs to be scaled with the total number of cases [but] we do have more confidence in the potential inflection in hospitalizations that our model indicates is happening now”.
That inflection was, in fact, unfolding when Kolanovic wrote those lines late last month. Below is a visual derived from official state data.
Incredibly, the mainstream financial media devoted almost no coverage to the half-dozen notes Kolanovic penned on COVID-19 throughout the crisis.
That, in hindsight, was a rather glaring editorial mistake, and one that cannot easily be dismissed as accidental.
For those unaware, the financial media goes out of its way to get Kolanovic’s name in a headline whenever they get an opportunity. It’s good for business, where “business” means web traffic and monetization of that traffic.
I am certain that most, if not all, of the notes cited above made their way at least to Bloomberg, and probably to CNBC and other mainstream outlets as well. One could scarcely conjure a more click-friendly headline than one which includes “coronavirus” and “Marko Kolanovic”.
The reality appears to be that for whatever reason, these notes (and the predictions they contained) were deliberately left out of the coverage mix, most glaringly on Bloomberg, which has a long history of covering his every utterance.
When Bloomberg finally got around to mentioning the latest note from Kolanovic, out earlier this week, it was the subject of a somewhat derisive blog post which appeared on the terminal and seemed to mock Marko for suggesting that sunshine interacts with the virus. Here’s an excerpt from something called “JPMorgan looks toward the sun to slow virus” (you can access it on the terminal – the condescending tone is unmistakable):
Given the still-fuzzy nature of global data surrounding the virus, any predictions like this should be taken with a grain of salt. However, the fundamental data is even fuzzier, ergo this is a market that can be influenced by themes and narratives. And this one is bound to pick up steam. One more reason to root for some sunny days this spring, anyway.
CNBC resorted to a similar tactic running a patently misleading headline which suggested that Kolanovic was basing his call for US equities to reclaim record highs early next year on “sunshine killing [the] virus”.
That is not, in fact, what Kolanovic said – at all.
The note (which you can read a thorough assessment of here) contains a lengthy rationale for why equities may fully recover by the first half of 2021. In the section devoted to that market call, there is not a single mention of “sunshine”. That discussion (of sunlight and the virus) is in an entirely separate section of the note.
In addition to the derisive blog post mentioned above, Bloomberg (in the “MLIV” blog section on the terminal), ran at least two pieces over the past several weeks which appeared to obliquely reference Kolanovic. Here’s one such excerpt which appeared on the terminal at 7:27 AM ET on April 17:
I’m not gonna lie to you: the market reaction to this reopening play is bemusing and confusing. Spooz are trading like it’s all back to normal in a matter of weeks, and while the opinions of armchair epidemiologists may not be worth much, actual ones are suggesting that the timetable is too aggressive.
As it turns out, this particular “armchair epidemiologist” predicted coronavirus outcomes in New York with the same remarkable precision as some of the market twists and turns he’s called over the years (including, by the way, his early January 2019 call that even after the worst December for equities since the Great Depression, the S&P would be back at all-time highs within “months”, which it was).
Your guess is as good as mine (and probably Marko’s) as to why these half-dozen notes were left out of the mainstream financial media’s coverage mix despite reporters’ long-standing habit of using his name to generate traffic at each and every opportunity.
Unfortunately, this is a case where their loss is also society’s loss. Because had these predictions been publicized earlier, they would have helped inform investors and, more importantly, policymakers, looking for answers about the likely timeframe on the epidemic. I covered them in these pages, but that’s clearly not adequate when it comes to getting the word out on something this important.
To the good folks at Bloomberg (and there are a lot of good folks there) and also to the mainstream financial media more generally, I would simply remind you that Kolanovic isn’t just “the man who moves markets”, as CNBC famously dubbed him. He’s also a PhD in theoretical high-energy physics, and boasts top-cited research papers in quantum gravity (you can look those up on Google Scholar, if you’re so inclined).
In other words, Kolanovic is a rocket scientist – almost literally. Bloomberg has a lot of former traders writing for them, especially on the terminal, but as far as I know, they’re short on rocket scientists.
So, if your question is why you should listen to Kolanovic when it comes to something like, for example, modeling epidemics, I would refer you to one of the many famous exchanges from the film Margin Call, a fictionalized retelling of the 2008 meltdown in mortgage-backed securities.
The exchange is between two executives at an investment bank (Robertson and Cohen), and an analyst (Sullivan) who figures out the bank’s VaR model no longer makes sense.
Sarah Robertson : What’s your background?
Peter Sullivan : My background?
Sarah Robertson : Your CV.
Peter Sullivan : I’ve been with the firm for two and a half years working with Eric that whole time, but I hold a doctorate in engineering, speciality in propulsion, from MIT, with a Bachelor’s from Penn.
Jared Cohen : What is a ‘specialty in propulsion,’ exactly?
Peter Sullivan : My thesis was a study in the ways that friction ratios affect steering outcomes in aeronautical use under reduced gravity loads.
Jared Cohen : So, you’re a rocket scientist.
Peter Sullivan : I was, yeah.
Jared Cohen : Interesting. How did you end up here?
Peter Sullivan : Well, it’s all just numbers really. Just changing what you’re adding up.
You must be logged in to post a comment.