September 12th, 2025 | Should members of Congress be allowed to trade stocks? Risks to Nvidia stock that you may not realize! Understanding AI and why it's becoming more expensive, 529 Plans & More

Should members of Congress be allowed to trade stocks?

I recently saw there was a bipartisan bill presented in the House that would ban lawmakers from trading individual stocks. I feel like we have been hearing about this for years, and according to NPR, “For more than a decade, a series of bills have been proposed to address such trades, but differences about the details and a lack of support from top congressional leaders stalled past reform efforts.” The question is, will this time be different? The bill made me curious though about how active congress was when it came to trading and let’s just say I couldn’t believe the numbers! In 2022 154 members of Congress made 14,752 trades, in 2023 118 members made 11,491 trades, in 2024 113 members made 9,261 trades, and through July of 2025 108 members made 7,810 trades. That is a crazy amount of activity and I’m not sure how they even have time for that. Their returns were also quite impressive with Democrats producing an average return of 31.1% in 2024 and Republicans producing an average return of 26.1%. For reference, the S&P 500 was up 23.3%. The numbers were quite staggering when you look at the individual performance of some of these politicians. In 2024, Rep. David Rouzer (R-NC) was up 149.0%, Rep. Debbie Wasserman Schultz (D-FL) was up 142.3%, Sen. Ron Wyden (D-OR) was up 123.8%, Rep. Roger Williams (R-TX) was up 111.2% and Rep. Nancy Pelosi (D-CA) rounded out the top ten with 70.9% return. These are hedge funds that are beating returns in several cases! Personally, I think it is ridiculous that politicians can trade individual stocks, and I hope there is finally action in Congress that ends it!

 

There are risks to Nvidia stock that you may not realize!

There is no denying what Nvidia has done has been extremely impressive, but one major problem with the company is the revenue is extremely concentrated. Their top customers made up 23% of total revenue in the recent quarter, which was up from 14% in the same quarter last year. Their second largest customer made up 16% of total revenue, which was up from 11% in the same quarter last year. Sales to four other customers contributed 14%, 11%,11%, and 10% of revenue respectively. This means that six customers accounted for 85% of Nvidia’s total sales. My concern is what if one of them drops out of the AI arms race or if a few of them pull back spending, that could really slow Nvidia’s business. I also believe that China is a risk to Nvidia. While sales have been hindered in the country due to political constraints, I believe many investors are looking to China as an area of potential growth for the company. All I can say to that, is do you really think the Chinese government wants Chinese companies using Nvidia chips? It was reported that Alibaba has recently developed an advanced chip, and I’d assume Huawei and other Chinese companies are racing to compete against Nvidia. While Nvidia stock essentially just keeps climbing, it’s important to realize there are several risks that could take the stock down! 

 

Understanding more about AI and why it's becoming more expensive We are no expert on artificial intelligence, but we have learned that while AI has gotten smarter it has also gotten more expensive. It is now broken down into a unit of AI which is known as a token and while the price of tokens continues to drop, the number of tokens needed to accomplish a task is increasing dramatically. There are two basic attributes to AI, one is called training, and the other is AI inference. The increase in cost is coming from the training side that has to use large models and demands even more costly processing. AI applications are using so-called reasoning and new forms of AI double check queries on their answers, which may include scanning the entire Web. Sometimes they write their own programs to calculate things all before releasing an answer that may only be a short sentence. Delivering meaningful and better responses takes a lot more tokens to complete that process. Looking at examples, basic chatbot Q&A requires 50 to 500 tokens. Short document summaries can be used anywhere from 200 tokens to 6000 tokens. Lawyers and paralegals who use legal document analysis require 5,000 to 250,000 tokens. If one is trying to do multi-step agent workflows, well now you’re looking at 100,000 to over 1 million tokens. Please understand when we talk tokens we’re not talking about anything that has to do with cryptocurrencies, and this is a different token pertaining to AI. Some big companies are spending $100 billion a year or more to create cutting-edge AI models and building out their infrastructure. However, for all that investment there needs to be a return on investment, and businesses and individuals will eventually have to pay more for artificial intelligence. The CFO of Open AI said last October that 75% of the company’s revenue comes from your average person paying $