Potatoes, Chernobyl, and AI: Exploring Market Reactions with What-If Scenarios
Henrique Heiss Zappe
Internationalist, Project Manager, Data Analyst - Seeking and spreading knowledge
What do potatoes have to do with Chernobyl? At first glance, probably nothing. But renowned author Michael Lewis tells us a different story. In one of his best-sellers, Liar’s Poker (1989), Lewis dives deep into the daily life of traders, salesmen, and the management team inside Salomon Brothers during the '80s. Among these stories, there is one character, in particular, who caught my attention. Today, my goal is to understand what principles drive this character and to see if new groundbreaking technology in Artificial Intelligence (AI) is able to “think” like this person does through hypothetical what-if scenarios. After all, this person knows what potatoes have to do with Chernobyl.
The person under the spotlight goes by a pseudonym he himself chose: Alexander. In Michael’s words, “Alexander was unique, the closest thing I met to a master of the markets.” Every day, Alexander would be called by managing directors worldwide who wanted advice on how to invest their money. Additionally, his specialty was how fast he could react to financial distress in the global market. At first, it seems like Alexander was a natural, capable of reading the reactions of the herd without the help of anybody. Yet, like every other successful person, he had a few principles that guided his approach to new information in the market. These were the contrarian and second-effect principles.
The contrarian principle states:
“If everybody is doing something, do the opposite.”
Haven’t we heard this in every Wall Street book? Although some might argue this is now obvious, I’d argue that doing this in practice is far harder than it seems. You have to have the guts to hold a position for an indefinite time. The worst part is yet to come: you will have nobody to blame if you are wrong. It will be on you. The truth is that being a contrarian demands a lot of confidence—the kind that comes from reliable information. Nevertheless, Alexander was a master at this. For instance, he profited from the Farm Credit System crisis in the '80s. Even though it was common knowledge that the U.S. government would not let the agency go bankrupt, people were selling its bonds like there was no tomorrow. For Alexander, those moments of panic were golden opportunities. The next time you find yourself navigating the seas of the stock market, try the conscious exercise of seeing through the lens of those taking the opposite side of the trade.
The second-effect principle is clever and brilliant. In its raw form, it’s a game of scenarios. When the Chernobyl tragedy occurred, Alexander called Lewis and said: “Buy potatoes,” then hung up. Would that be your first reaction in such a tragedy? Probably not. The market’s initial reaction was to heavily sell stocks related to nuclear power. Alexander, however, bought a significant amount of potatoes within minutes of the event. His logic was simple: a fallout cloud in Europe would compromise farms, causing Europeans to look for American potatoes as a substitute. He also bought a huge amount of oil, predicting that a reduced supply of nuclear power would increase oil demand. The lesson here is:
“When a major event happens, forget where everyone else is focusing and seek secondary or even tertiary effects.”
In my opinion, this secondary effect principle inherently contains a bit of the contrarian mindset. In brief, look where nobody else is looking, have confidence in your instincts, and you’ll be rewarded.
Michael Lewis claims that Chernobyl and oil are straightforward examples. Maybe for him. For those new to the market—or for those who have had bad experiences—these ideas may not be so simple. However, there is a way to prepare for future events. In his opinion, challenges or periods of distress can be played out through a game called "What-If." The game involves identifying problems and considering what might happen next. An oversimplified example might be: “What if a plate falls on the ground?” The obvious answer: “It’ll break!” But what else might happen? Your wife might get angry; your dog could start barking; the baby might cry; your feet might get cut; and so on. I think you get the gist.
Now that we understand Alexander’s thought process—the contrarian and second-effect principles—we can proceed to update these thoughts from the '80s with today’s technology. I want to put AI to the test in the What-If game.
What-If Scenarios
I will prompt ChatGPT with enough context to understand the game and then ask it to play the What-If game two times, using different scenarios, but driven by Alexander's principles.
Here are the scenarios I chose for this exercise:
Scenario 1: What If China Invades Taiwan?
This scenario concerns international affairs. It’s a topic that regularly appears in headlines worldwide. According to ChatGPT, here are the results:
First Reaction: “Many investors would likely pull out of Asian markets, fearing instability and the geopolitical ripple effect. Defense stocks, especially U.S.-based ones, would surge due to expected military involvement, while technology stocks could fall given Taiwan’s role in the semiconductor industry.”
领英推荐
Contrarian Opportunity: “Go against the panic-driven sell-off in specific Asian markets not directly tied to the conflict. Southeast Asian countries like Vietnam, Thailand, or Malaysia might experience initial pullbacks due to regional fears, but these economies could benefit as manufacturers relocate to avoid China-related risks. Buying in these markets post-panic might offer long-term gains as these countries gain new business.”
Most investors would probably pull out of Asian markets—especially China—for obvious reasons. A surge in defense stocks is also expected, given probable U.S. involvement. I appreciated the contrarian opportunity identified by ChatGPT. It recognized that heavy selling would create opportunities in nearby countries like Vietnam, Thailand, and Malaysia.
For secondary effects, I initially found them uncreative, including oil changes, logistics re-routing, and semiconductor shortages. However, after pressing ChatGPT for more out-of-the-box ideas, it suggested:
With deeper inquiry, AI can spotlight opportunities we might overlook. Future investigations could assess whether these effects would indeed stem from such an event.
Scenario 2: What If We Have Another Pandemic Outbreak?
The COVID-19 pandemic was a terrible moment at the start of this decade, and we may never know how quickly another outbreak could occur. Procedures were created that would certainly be reused. If it happened again, how would the market react, and what would be a contrarian move?
Mainstream Move: “Most people would expect a repeat of the previous pandemic—investors would rush to pharmaceuticals, vaccine producers, digital communication platforms, and health stocks. Hospitality, travel, and retail would take an immediate hit.”
Contrarian Opportunity: “If the market runs to big pharmaceuticals, consider small-cap health-tech startups instead. Governments and people are now much more attuned to rapid-response health measures. Small, agile companies developing point-of-care testing, wearable health monitors, or remote diagnostic tools could benefit—these companies might have innovative solutions that are not yet under the mainstream radar.”
In the middle of another pandemic, healthcare is naturally the focus. ChatGPT’s contrarian approach here is well-founded in historical trends, highlighting small health-tech startups rather than big pharmaceuticals.
Continuing with secondary and tertiary effects:
This response is quite rich, spanning different industries, all with realistic links to a new pandemic scenario. The connection between industries like pet health, urban agriculture, and therapy services is particularly compelling.
Conclusion and Next Steps
In summary, the “contrarian” and “second-effect” principles are crucial for those venturing into the market. Alexander, a master of the market, adhered to them and could see clearly through market panic during big events. Today, AI can assist us in predicting likely (and perhaps less obvious) outcomes of major events. It's up to us to determine if AI's predictions are hallucinations or if they hold historical credibility. I plan to explore this topic further in my next article.
Do these principles help you make better market decisions? Do you think AI was able to identify real opportunities here? Why?
If you're an expert and would like to discuss this article or any related topics, please send me a message on LinkedIn or email me on: [email protected]