At this point, there is no point in debating the utter dominance of Nvidia (NVDA). After a rough start in 2025, NVDA stock has found its rhythm once again, returning stakeholders more than 35%. I wouldn’t call it life-changing performance per se. However, the freight train appears unstoppable.
What’s more, Nvidia is scheduled to release its earnings report on Aug. 27 after the market close. Undoubtedly, it’s going to be the disclosure that everyone on the Street will anxiously consume. Obviously, Nvidia represents one of the elite names in artificial intelligence. It’s also inking major deals for its most advanced AI chips. Further, analysts overwhelmingly rate NVDA stock as a consensus Strong Buy.
The list of fundamental strengths goes on and on. So, despite what many might consider sky-high multiples, the argument is that NVDA stock is undervalued relative to its true potential. However, this argument presupposes that the market hasn’t already priced in all the good stuff into NVDA. It would be an extraordinary claim that financial publication readers recognize Nvidia’s true value but the rest of the market does not.
That doesn’t mean that NVDA stock won’t swing higher — but for short-term traders and even long-term investors, there might be better ideas available.
Let’s consider the first-order principles of market demand. In the trailing 10 weeks (including this one), the market voted to buy NVDA stock seven times and sell three times. During this period, NVDA enjoyed an upward trajectory. For brevity, we can label this sequence as 7-3-U.
At first, it might sound strange to compress NVDA’s price action into a simple binary code. But now we have a falsifiable sequence by which we can extract forward probabilities through the study of past analogs. Long story short, when the 7-3-U sequence flashes, there’s only a 52.17% chance that the following week’s price action will rise.
The issue that I have is that the baseline probability is 59.25%. This is just descriptive math. I’m not saying I hate NVDA stock. I’m just saying that this particular hand at the blackjack table — the 7-3-U sequence — is unfavorable to the bullish speculator.
As an alternative, if you want to take a bet on a semiconductor specialist, investors may consider ChipMOS Technologies (IMOS).
Based in Taiwan, ChipMOS Technologies specializes in semiconductor backend services — notably testing, assembly, packaging, bumping and wafer-level chip scale packaging solutions. Its services cater to a broad range of integrated circuits, including memory, mixed-signal, display driver ICs and ASICs.
Of course, Nvidia and ChipMOS are not related, other than the fact that both enterprises are under the broad semiconductor umbrella. Nvidia is a chipmaker, specifically a fabless semiconductor company. In contrast, ChipMOS is a semiconductor backend services provider: it takes chips that have already been fabricated and tests and packages them, along with other assembly services.
But the point of bringing up ChipMOS is not the business itself but the potential discount in IMOS stock. During the midweek session, IMOS stock dropped nearly 5%. Since the start of the year, it’s down roughly 12%. In fact, I selected IMOS in part because it was included in Barchart’s screener Three Day Losers.
Obviously, finding losers is easy business. Finding ones that can recover is the hard part. However, in the case of IMOS stock, I believe a comeback is possible based on a key quantitative signal.
Using the discrete-state analysis mentioned earlier for NVDA, in the trailing 10 weeks, the market voted to buy IMOS stock three times and sell seven times. During this period, the security suffered a downward trajectory. We can call this a 3-7-D sequence:
Now that we have a falsifiable signal, we can look back to the end of the dataset (January 2019) to discover how the market responded to the sequence. The name of the game is to only act when the signal can reliably beat the baseline probability or what is effectively our null hypothesis — the assumption of no mispricing.
As it turns out, the chance that a long position in IMOS stock may rise on any given week is exactly 50%, a coin toss. However, our alternative hypothesis is that, because of the 3-7-D sequence, a favorable mispricing exists. Indeed, past analogs suggest that the odds of upside stand at 60.87%.
Assuming the positive pathway, over the next three weeks, IMOS stock could potentially hit $17.41, making it an intriguing buy-the-dip prospect.
Anytime you’re dealing with a market signal, you want to see if it’s distinguishable from random noise. Therefore, I ran a one-tailed binomial test on the 3-7-D sequence, which spits out a p-value of 0.2024. Essentially, this means that there’s a 20.24% chance that the implications of the signal could materialize randomly as opposed to intentionally.
Admittedly, that’s a bit on the high side. However, you must also consider that the stock market is an open and entropic system. In other words, exogenous factors can easily enter the paradigm and disrupt the market’s trajectory. There’s no model that can consistently predict such impacts.
At the same time, the math that I’m running is descriptive. Stated differently, the forecasts are based on what has happened in like-to-like conditions. In traditional Western models, the math is prescriptive, which focuses on what should happen.
I believe the descriptive approach and forward extrapolation are much more faithful to actual market dynamics. But bottom line, you should consider the model with which you’re most comfortable.
On the date of publication, Josh Enomoto did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. This article was originally published on Barchart.com