Archive
-
How a Law Becomes a Rule
Agencies, not the president or Congress, set most of the rules governing the US. But citizens have a say.
-
Why India's unconventional foreign exchange policies won't work
Understanding India's unconventional monetary policy through T-accounts
-
Oil prices, not safe havens, are driving the dollar's rally
The dollar's surge has more to do with the price of oil than the flight to safety
-
Why making it easier to short a currency can weaken it
To slow the RMB's appreciation, the PBoC is making it easier to short the currency. Looking at the role of dealers hedging inventory risk shows how this rule change may work.
-
Private credit took the loans, stablecoins threaten the funding
Banks are already facing competition for loans. Yield-bearing stablecoins will make funding these loans more expensive.
-
Does the Fed manipulate rates in election years?
Using implied forward rates to back out expected interest rates in election years shows the bond market does not expect Fed manipulation.
-
External Validity
Understanding when model retraining won't work
- Week 14 NFL Predictions, Daggy vs Kalshi
-
Losing money at video poker, with data science!
Comparing A/B testing and bandit algorithms to find the best slot machine.
-
Mortgage rates under various economic scenarios
Using Daggy to analyze how economic scenarios will impact mortgage rates
-
Geopolitical Shocks and Stocks
Why geopolitical shocks impact the stock market much less than we think
-
Is July really a bad month for stocks?
Someone at lunch claimed July is bad for stocks. With Daggy, we built 8 statistical models to show this isn't the case and revealed why raw data deceives.
-
LLMs represent a win for utility theory
AI's success shows we don't need to encode heuristics and biases into agent based models
-
Innovation, knowledge workers and collective intelligence
What John Kay's complexity approach in The Corporation in the 21st Century says about AI taking our jobs.
- What political science teaches us about AI SEO
-
When absence of evidence is evidence of absence
Bayes' rule shows us when a lack of evidence is evidence of absence
-
How Text Data Can Improve (or Worsen) Business Decisions
Teams often want to answer questions like "Does price impact our satisfaction scores?" But hidden variables can confound these relationships. This post explains how text data from reviews and feedback offers both promising solutions and potential pitfalls for uncovering true business insights
- Shapley Values for Interpretable ML
-
Can Biden Make Bibi Balk?
Applying game theory to Biden's attempt to halt Netanyahu's invasion Rafah
-
Sources of Polling Error
This post provides an overview of polling and explores the margin of error in a poll.