adverse selection in the option job market
Friends,
Today I’m going to both be a student and speaker at Ricki Heicklen’s Quantitative Bootcamp in Berkeley. Ricki is a Jane Street alum who got on my radar when I heard her interview with Patrick Mackenzie.
I summarized the convo in A Jane Street Alum Teaches Trading but it’s still a long article. The interview is dense with insight. Very rare for this stuff to be presented in such an accessible way in public.
Early in the interview Patrick asks Ricki to complete this sentence:
“I’m going to impart a bit of information upon you to get you ready for understanding the US equities markets…” – if you only had one sentence, what is that?
Ricki, without hesitation:
The number one sentence for purposes of trading, in general, is to think about adverse selection.
The questionnaire she give to incoming students is loaded with tricky examples of adverse selection hiding in seemingly innocuous scenarios. You can get some flavor of this in her post Toward a Broader Conception of Adverse Selection.
💡Aside: Polling has been a popular topic recently for obvious reasons. Adverse selection is a sibling of sampling bias. It’s the mother of monkeywrenches in statistics. This back and forth between Ben Orlin and Jim O’Shaughnessy has several fun, yet profound, examples of sampling bias.
Mathematician Ben Orlin on Infinite Loops podcast audio & transcript (6 min)
Fellow Jane Street alum, Agustin Lebron, has also emphasized the significance of adverse selection. In his exceptional book, Law of Trading (my notes), chapter 1 is about motivation. Chapter 2 is, you guessed it, Adverse Selection.
[The obsession here is not misplaced. SIG or any other market maker is going to dwell on this because markets are not kindergarten — your counterparty wants to make money by disagreeing with you so understanding if they are safe to trade against is the a primary objective.]
The chapter includes several great examples of adverse selection in financial markets, but as every chapter does, it closes with practical applications that are pertinent to anyone not just traders. I’ll use this extended excerpt about adverse selection in the labor market to setup today’s discussion (emphasis mine):
Adverse selection in the job market appears on the side of both the employer and the prospective employee. Employees are looking for the most attractive job they can find, while employers are looking for the best candidate for the job.
Prospective employees are subject to various selection pressures, not all of them adverse:
Given that the company is looking to hire employees, things can’t be going all that badly for the firm. This could have a positive selection effect, for once. Nevertheless, maybe the company is hiring because people are leaving and they’re having trouble finding hires. It’s a double-edged sword.
The job description is likely to be embellished, at least a little bit, especially for less-attractive jobs.
The interviewers in a company are typically better-than-average employees since they’re the ones the company chooses to represent them to potential hires. Thus, the applicant sees a rosy picture of the quality of her future co-workers.
Employees typically only seek out new jobs every few years or so. As a result, their skill at job finding is lower than the companies’ skill at candidate finding. This asymmetry of skill and knowledge works in the companies’ favor.
Employers, however, suffer significantly greater adverse selection, since in the end the employee is the one making the final decision of either accepting or rejecting a job offer [Kris: the candidate holds the last option — it’s like the asymmetry in backgammon with the doubling cube — you must have significant edge to offer it, but you only need to have a 25% chance of winning to accept it. Btw, this was an interview question I had back in 1999]
When hiring people with prior experience, the applicant pool skews in the direction of lower-quality workers. This is because good workers will be preferentially incentivized to remain with their current employers. On average, therefore, people with prior experience looking for jobs aren’t as good as those who have jobs (Greenwald, 1986).
The process of interviewing to decide whom to hire is an imperfect one. Companies will therefore not be able to sufficiently distinguish between good and bad workers, meaning they will tend to squeeze offered wages toward the average. This is advantageous for less competent workers, who, for the most part, will be the ones looking for work, and disadvantageous for good workers, who will be driven away from the job market by the lower-than-deserved wages on offer. [Kris: this is the so-called “lemon problem” in the used car market]
Potential employees will select the best offer from all the ones available to them. Good workers will have a large pool of available offers and will pick the best one. If a given employer isn't universally known as the most desirable one, then it’s fair to say that a worker who accepts an employer’s job offer does so because she couldn’t get a better one elsewhere.
A reasonable blanket assumption is the employer faces a larger adverse selection problem than the candidate.
But what I want to talk about is how with experienced option traders I have seen this flipped. Well, maybe not flipped, that’s hard to say but I can pinpoint a somewhat narrow but substantial area of adverse selection that I’d expect was more prominent in the past 5 years.
Whenever I have a call with a trader evaluating an opportunity I warn them about this because I’ve seen it time and time again.
No math or charts today. This is a topic for experienced professional option traders although I’d bet it could be generalized faithfully to anyone with valuable experience who is being sought after. You can make the adjustments for your own field.
I’m going to be very direct since you are not a learner or novice. If you’re at this point you understand the angles. You are already damn good at what you do. The failure mode here has nothing to do with your competence as an option trader or risk taker.