Leland Bybee @Leland_Bybee Twitter profile | Pikagi

Pikagi

Leland Bybee

@Leland_Bybee

639

Followers

1,359

Following

11

Media

43

Statuses

Incoming finance AP @ChicagoBooth . PhD candidate in financial economics @YaleSOM .

https://t.co/Bcq056Rnw2

Joined June 2010

Don't wanna be here? Send us removal request.

Pinned Tweet

@Leland_Bybee

Leland Bybee

6 months

Couldn't be happier to announce that I have survived the job market and will be joining @ChicagoBooth as a finance AP this summer! It was amazing to meet so many awesome people on the market and many thanks to the fantastic support from everyone at @YaleSOM !

20

4

260

Last Seen Profiles

@kaihimo_no1

@bokeplokalmalam

@CGB_FR

@elfloppeur

@ClintonTopps

@SamEwen

@2JUJU_

@uthyimbiriaiwoy

@SisyphosGn

@SisyphosGn

@himena__u

@zaralla1

@GO_AHEADPEOPLE

@Sinaco_O

@1lfi_

@adhdkuroo

@mistedoll

@t2ndsun

@AthleteStory

@Junior_Optimist

@gayamman22

@S2_Sarasa

@TimExcluded

@NF4331

@BuzzardMustard

@netterdan

@liggonfit

@the_luoya

@SparkSigns

@bokeplokalmalam

@livlandisjets

@LymelightTO

@Arz_XqX

@muhammadas90242

@ALighthouseCat

@itsmejuyy

@Leland_Bybee

Leland Bybee

10 months

Thanks Arpit! What better time to post about my AI-related JMP, "The Ghost in the Machine: Generating Beliefs with Large Language Models", a 🧵: 🤖🧠🤖🧠🤖🧠🤖

4

30

183

@Leland_Bybee

Leland Bybee

10 months

This sentiment series captures the over/under enthusiasm we’d want from a measure of systematic errors: it correlates strongly with Shiller’s PE ratio and negatively predicts future returns.

Tweet media one

1

2

11

@Leland_Bybee

Leland Bybee

10 months

There’s a lot more in the paper, which can be found here: Welcome any and all thoughts!

1

0

10

@Leland_Bybee

Leland Bybee

10 months

How do I generate expectations? I provide an LLM with news articles and ask what it expects will happen to various economic outcomes (S&P, indpro, CPI...) based on each piece of news. I then aggregate these article-level responses to form a time series of generated beliefs.

Tweet media one

1

1

8

@Leland_Bybee

Leland Bybee

10 months

So what can we do with generated expectations? For one thing, greatly expand the sample of expectations data – I use news text back to 1900 to build expectations for a broad set of variables and then extract a measure of the systematic errors or “sentiment” over that sample:

Tweet media one

1

1

8

@Leland_Bybee

Leland Bybee

10 months

Behavioral theories are difficult to evaluate in asset pricing because we don’t directly observe investor beliefs. Nowhere does this matter more than for theories of bubbles. Can we use generated expectation to help? Yes!

Tweet media one

1

1

7

@Leland_Bybee

Leland Bybee

10 months

LLMs implicitly react to information in a similar way to humans (including all our biases and errors). If I could provide an LLM with a textual representation of the state of the world at a point in time I could use that LLM to generate “human-like” expectations.

Tweet media one

1

1

6

@Leland_Bybee

Leland Bybee

10 months

Importantly, generated expectations also exhibit many of the same deviations from full-information rational expectations that have been documented in the literature. Some summaries of those deviations for return expectations:

Tweet media one

1

1

6

@Leland_Bybee

Leland Bybee

10 months

Expectations are central to econ, but aren't directly observable. Rational expectations (brilliantly) solves this but we have mountains of evidence against RE. Without RE we're in a "wilderness of alternative expectations" - I propose a new method to navigate this wilderness…

Tweet media one

1

1

6

@Leland_Bybee

Leland Bybee

10 months

@cameron_pfiffer @shoshievass @yizhoujin This is sometimes done with Gibbs sampling for latent Dirichlet allocation -- basically do updates in parallel on each chunk, then recombine each step -- works quite well and can scale to very large datasets.

1

0

4

@Leland_Bybee

Leland Bybee

10 months

I also find evidence for return extrapolation as a key driver of these bubbles:

Tweet media one

1

1

6

@Leland_Bybee

Leland Bybee

10 months

I find that an asset’s exposure to sentiment can be used to predict bubbles ex-ante -- heavily exposed assets that experience a run-up are much more likely to crash and experience negative future returns.

Tweet media one

1

1

5

@Leland_Bybee

Leland Bybee

10 months

How “human-like” are generated expectations? To test this, I compare my generated expectations to various surveys. Here’s the generated expectations for the S&P 500 against two commonly used return surveys:

Tweet media one

1

1

5

@Leland_Bybee

Leland Bybee

10 months

This strong connection with existing surveys holds for a wide range of variables. Here are correlations between generated expectations and the survey of professional forecasters for a number of macro series:

Tweet media one

1

1

5

@Leland_Bybee

Leland Bybee

8 months

@momin_rayhan @alejandroll10 @achenfinance This papers is pretty related to what your suggesting:

1

0

2

@Leland_Bybee

Leland Bybee

1 year

@AnthonyLeeZhang Not quite that but:

Tweet card media

Surveying Generative AI's Economic Expectations

I introduce a survey of economic expectations formed by querying a large language model (LLM)'s expectations of various financial and macroeconomic variables based on a sample of news articles...

0

1

2

@Leland_Bybee

Leland Bybee

2 years

@momin_rayhan This is intuition is essentially how you get simulated annealing:

Simulated annealing - Wikipedia

en.m.wikipedia.org

1

0

2

@Leland_Bybee

Leland Bybee

1 year

@paulgp Thanks @paulgp ! Planning to do a thread on the results soon!

0

0

2

@Leland_Bybee

Leland Bybee

6 months

@DhruvChandAgga1 @ChicagoBooth @YaleSOM Thanks and definitely looking forward to it!

0

0

0

@Leland_Bybee

Leland Bybee

6 months

@stevehouf @ChicagoBooth @YaleSOM Thanks!

0

0

0

@Leland_Bybee

Leland Bybee

6 months

@SashaIndarte @ChicagoBooth @YaleSOM Thanks!

0

0

1

@Leland_Bybee

Leland Bybee

9 months

@arpitrage Somewhat related but there is evidence for a US/Euro bias in GPT-3.5/4:

1

0

0

@Leland_Bybee

Leland Bybee

6 months

@m_aragoneses @gk_ben @ChicagoBooth @YaleSOM Thanks!

0

0

1

@Leland_Bybee

Leland Bybee

6 months

@TaniaBabina @ChicagoBooth @YaleSOM Thanks!

0

0

1

@Leland_Bybee

Leland Bybee

6 months

@alejandroll10 @ChicagoBooth @YaleSOM Thanks!

0

0

1

@Leland_Bybee

Leland Bybee

9 months

@arpitrage Totally

0

0

0

@Leland_Bybee

Leland Bybee

6 months

@MissChangHe @ChicagoBooth @YaleSOM Thanks!

0

0

1

@Leland_Bybee

Leland Bybee

6 months

@_brit_lewis @ChicagoBooth @YaleSOM Thanks!

0

0

1

@Leland_Bybee

Leland Bybee

6 months

@alizarif @ChicagoBooth @YaleSOM Thanks!

0

0

1