Does partisanship shape investor beliefs?

We use party-identifying language – like “Liberal Media” and “MAGA”– to identify Republican users on the investor social platform StockTwits. Using a difference-in-difference design, we find that the beliefs of partisan Republicans about equities remain relatively unfazed during the COVID-19 pandemic, while other users become considerably more pessimistic. In cross-sectional tests, we find Republicans become relatively more optimistic about stocks that suffered the most from COVID-19, but more pessimistic about Chinese stocks. Finally, stocks with the greatest partisan disagreement on StockTwits have significantly more trading in the broader market, which explains 20% of the increase in stock turnover during the pandemic.

Here is the full piece by J. Anthony Cookson, Joseph Engelberg, and William Mullins, via the excellent Samir Varma.


How do we infer the direction of causality? Opinions relevant to some equity prices likely dispose one more to one side politically. If it’s March and you think COVID-19 is gonna be particularly horrible, you’re probably both a pessimist about equities and not sympathetic to Republican politicians for the same reason.

I think that was exactly the point of the study. The causality is found in the price of the stocks: pessimistic (Democratic Party / Chinese) stocks fall further than optimistic (Republican party) stocks.

Coupled with the previous TC link about stock market investors becoming more conservative over time, and, with the observation about the "cult of equities" since at least the mid-1980s (nearly 40 years ago!), can this explain why stock markets "always go up"? Can it be that stocks are now seen like real estate, in that they "never go down"?

I think perhaps so, and it proves the Keynes observation that demand is what matters in a modern economy, not supply. If people think an economy is strong, it becomes strong. Perhaps this explains why African economies never seem to take off? Permanent pessimism leads to a vicious cycle, while in advanced economies optimism leads to a virtuous cycle. It explains why places with no real natural advantage, like NZ, SA, CHE, arguably even Australia (outside of the mining towns), Austria (does tourism, e.g. alpines and opera, prop them up, like it does GRE?) seem to do well: people 'will' it so. It's a different narrative about advanced economies from the usual 'white people win'. Or perhaps another perspective, maybe white folk are more optimistic due to their historical successful oppression of non-white folk?

By email I've been trying to nudge TC and some other famous economists along this direction, to do a post on this, that people can will a recession or will an expansion, but they won't listen or reply to me. So you read it here first. Expect a PhD thesis or three to follow from somebody aping this theme. Also dovetails with the Paul Krugman thesis about trading in successful modern economies (first in time becomes first in right, initial conditions matter).

The causality is simple. We created out of thin air trillions of dollars in the past month. It has to go somewhere. The dollar is the best currency of the bunch. Stocks looks relatively better than bonds and real estate. So to the US stock market it goes. A poll earlier that said 80% think the country is out of control. Printing money is the current technocratic solution to the pessimism out there. Will a stock market hitting near all time highs translate into real prosperity or will the paper gains driven by printing go out in a flash crash?

@PGBrrr... yes your point is that "money illusion" is driving NGDP and ultimately affecting real GDP in a positive, good way, aka the Scott Sumner thesis, with the corresponding 'wealth effect' contributing as well. I don't buy that, but I'm in the minority.

Incidentally, great minds think alike, Sumner anticipated this post a few days ago by blogging 'real shocks are not that important', which is the flip side of what I posted above. See:

I’ve lived thru 4 “stocks never go down” phases. It’s not pretty.

I learned my hard lesson in 2001. I thought I knew better. I was wrong.

But even then, there are good buys.

Oh, Christ: You gotta pick stocks the way Charlie Darwin says to. Of course, he won't tell us beforehand. Nothing else matters.

Interesting. Could this be some sort of indirect futarchy, where we can analyze the returns of these stocks and tease out which sets of beliefs are correct?

Can we do away with these stories equivalent to "liberals prefer chocolate ice cream while conservatives like vanilla"? This is the social science version of Cosmopolitan's "50 ways to Tease a Man" headlines. I don't care how many graphs and equations are there, it is gossip material for technocrats.

Why not? If it's falsifiable, it's science, not metaphysics. Unlike the quantity theory of money, I might add.

"Why not? If it's falsifiable, it's science, not metaphysics."

Is it accurate?

"We use party-identifying language – like “Liberal Media” and “MAGA”"

How often does that language on this site reflect an actual Republican vs someone else? There are plenty of trolls who use the term MAGA. How high is the correlation?

One of authors here. Rather than let open-ended questions stand unanswered when there's a perfectly good answer in the paper, I'll issue a clarification to these questions because we thought about them.

Our abstract couldn't be 100 words if we included the full classification. We built the wordlist to flag posts as likely Republican, but we were worried about misclassification of the type you describe. So, we require that a user make *or like* at least 3 words with these keywords.

In addition, we read a bunch of posts that contain these keywords, not hard to get a good sense as there were about 31,000 of them. The vast majority are using these terms genuinely without a hint of irony. A few are trolls, but that misclassification would bias us against finding differences between the groups.

Further, the measure "validates" in the sense that our Republicans become more optimistic around the 2016 election and at key moments in the US-China trade war.

I'll not go into more detail, except to encourage you to read the paper if you have questions about the accuracy of our methodology. We might not have an answer to everything, but we tried to be thorough.

In general, we are happy to receive comments and suggestions, as this is a working paper that we're hoping to improve as we move toward publishing it.

Caveat: If you think of something useful for us to consider, feel free to email it, as I'm not going to monitor the discussion comments.

Author perspective: I agree that a paper that says "liberals prefer chocolate ice cream while conservatives like vanilla" would be uninteresting, but that's mostly because it is not very costly for liberals & conservatives to disagree about ice cream flavors.

By contrast, it is potentially quite costly for investors to firmly hold onto beliefs that are not informed by information if it guides their trading. Our trading results and other research suggest that these beliefs do matter for more than opinion polling.

FWIW: We don't take a stand in the paper about who is paying the cost of holding non-informative beliefs. It might be that Republicans have unwarranted optimism, or that non-Republicans have unwarranted pessimism, or a mix of both. Still, from the standpoint of thinking about models of financial markets, we think this evidence is important.

Liberalism is a mental disorder. It's not my opinion, it's every behavioral study.

America long ago became committed to rising stock prices (and the concomitant investment in stocks by owners of capital) as the path to prosperity (as opposed to investment in productive capital). Indeed, the Fed signaled early on that the Fed would do "whatever it takes" to support stock and bond prices. That the stock market has recouped almost all of the losses suffered early in the pandemic is incentive for investors to stick with stocks (as opposed to, for example, productive capital) both now and in the future. On the other hand, if de-globalization is the future, owners of capital (U.S. firms) will have little choice but to invest in productive capital here; if they don't how will U.S. firms replace all those profits that were generated from supply chains in China and elsewhere? These are interesting times.

This very rapid stock market collapse and recovery has been kind of satisfying because it has proven everyone wrong. People who are consistent buy-and-hold optimists (like me) are doing about as well as the prepper types who sold in February and aren’t back in yet. No one can really say they had some great foresight that the Sheeple lacked and are now retired on a beach somewhere unless they sold in late February and bought back in in late March, and I think you’ll be hard-pressed to find many people who did.

Now hopefully the stock market stays flat for a spell or all the preppers come back in. There will be fewer pitchforks if people are all having similar investment returns, unlike in 2008-09 when lots of people lost their money, cashed out at the bottom, and then were bitter for years.

There are some who say life itself is based on unrealistic optimism!

(I see that REITs are still way down, so maybe people aren't betting unrealistically on rent collection.)

REITs are going to have a rough go for a while. There are a lot of people wondering why they need to go back into an office. I have friends that have been told that they'll be working from home until the end of the year. That's a lot of time to establish new habits.

The biggest permanent effect of Covid will be the massive acceleration of the work at home trend, which was a small and growing thing. Now it's everyone who can.

Which will lead to a permanent income increase. It's a significant savings for many workers in commuting costs, opens up housing to workers in remote areas (which will put downward pressure on housing prices in built up areas, reduces stress associated with commuting and potentially reduces long term disease transmission.

Assuming: the work at home is long term and doesn't have any costly unforeseen side effects,

Retail REIT's are likely to be hit pretty hard.

Not just temporary delays in collecting rent, but it really accelerated bankruptcies of marginal retailers, may mean some Chapter 11's become Chapter 7's, and also probably will mean more store closures even by chains that don't file.

Big acceleration of trend toward online shopping seems to be the consensus.

Though, I do wonder - if we actually get more WFH, will that unexpectedly help physical retail? People presumably want to leave their house to do *something*.

"Investors" is a generous term for Stocktwits users.

Comments for this post are closed