This should have been a 3 part-er but…
When I was in my first job as a “trader trainee” in addition to the standard mechanics of doing the job like knowing bids, offers, sizing, options greeks etc. etc. etc., the senior traders also stressed certain mentalities and abstract concepts that can make or break your career.
One of those ideas is being “smarter than the other guy”.
This isn’t referring to SAT scores or the count of Wiley Finance textbooks on your the shelf. It refer to the amount of market moving information one trader has over another to make decisions.
Each trade has a buyer and a seller. Each side (theoretically) believes they are correct in their decision to buy or sell the security. Whatever their reasoning may be, over time having more accurate information than another will result in more correct and profitable decision making.
Now, outright illegal insider trading is the most obvious example of this asymmetry in play but plenty of legal, yet naïve “information gathering” and “analysis” exists too.
Parroting themes of AI supremacy, speculative crypto and quantum computing plays will make someone feel awfully smart and confident about their portfolio positions.
Enron: The Smartest Guys in the Room is a 2005 documentary examining the Enron Corporation, their energy trading scandals and ultimate collapse in 2001
But do these traders really understand the narrative of the company or market they follow? With equities back at all time highs, led again by the newly minted $4 trillion dollar NVIDIA and talk of AI supremacy, let’s look at some narratives and decide if they still hold up.
Did you know?
Many of the tech layoffs in Silicon Valley since 2023 correlates very closely in timing and departments to a change in tax law on R&D expenses….and NOT necessarily due to A.I. implementation as the CNBC narrative suggest.
The 2017 Tax Cuts and Jobs Act (TCJA) had a specific Section 174 on Amortization:
Previous Law: Before 2022, businesses could immediately deduct the full amount of their R&D expenses in the year they were incurred, referred to as "full expensing".
Change: The TCJA mandated that starting in 2022, businesses must capitalize (treat as an asset) and amortize (deduct over time) their R&D expenses(including employee salaries). Domestic R&D is amortized over 5 years, while foreign R&D is amortized over 15 years.
Impact: This change has significantly increased the tax burden on companies by delaying the tax benefits of investing in R&D or keeping R&D related staff on payroll. The hardest impacted tech companies also happen to all have AI products to pump, so making AI the culprit puts a nice marketing bow on the layoff press release.
Did you know?
The supply chain of an NVIDIA H100 AI chip consists of 1000s of companies and pieces. Any disruptions to the chain will delay the next-generation training models and leave any AI product suite with middling models for public use. The most advanced AI will be out of reach for most organizations except best-in-class innovators with the most capital to burn.
Example: KLARNA originally announced a move to AI customer support touting the productivity gains and cost improvements. Later, they announced a pause and even a reversal hiring back humans after customers simply grew frustrated with AI chatbots’ lack of common sense.
IMPACT: Some top players in tech industries will have the resources to apply bleeding edge AI for their internal purposes but beware mid-tier CEOs claiming a miracle in productivity gains. For the rest of society, any supply chain disruption in chips and we’ll be stuck with ChatGPT-3 caliber intelligence for most companies, leaving AI as an augmenting tech rather than a replacement of workers.
Did you know?
Breaking cryptographic security is the go-to narrative for usage of quantum computing with a few lab demos done as proof-of-concept. It’s said with AI’s help, encrypted assets like bitcoin wallets or secure encrypted communications would be at risk. However, the hardware it will take 250x-1000x times the current energy, sizing and investment to get there.
Reality: Shor’s Algorithm can efficiently factor large numbers and could extract a private key from a public key in ~2,500–10,000 logical qubits. Currently no known quantum computer has anywhere close to this scale. IBM’s white papers on their progress indicate they can handle < 10 logical cubits at the moment.
IMPACT: Unless a breakthrough happens, we are at least 10–20 years away from practical capability of a quantum computer. Any start-up or IPO in the quantum area will be decades away from profitability or even a viable product.
Interested in being on the Trader Dads Podcast in 2025? Shoot me an email! I’d love to have subscribers on to sit for a discussion
TRADERDADS MAILBAG
Thoughts? Questions? Comments?
Reach out! Maybe I’ll do a full post on the topic or as a Q&A
traderdads@substack.com
