When Nvidia (NVDA) releases its must-see earnings report on Wednesday night, it will be faced with this frustrating reality. Strong earnings reports haven’t been too kind to Nvidia stock lately! Nvidia shares have fallen in response to earnings in six of the last eight quarters, including the last four, according to Yahoo Finance AlphaSpace analysis. The reality is that the market is positioned for the company to release something great and for CEO Jensen Huang to appear super bullish on the earnings call. The market also knows there is minimal downside risk to Nvidia’s growing investment portfolio, given the rising valuations (see Anthropic (ANTH.PVT), for example) given to most private names in artificial intelligence. “Because many of the discussions around AI infrastructure spend/return on investment and credit risk are outside NVDA’s purview, we believe the numbers are more important than the narrative and coming off this call, we expect investors to gain greater confidence on a path to $15+ EPS in 2027E and $20 in 2028E, numbers that should keep the stock rising,” UBS analyst Tim Arcuri said in a note. Considering that Nvidia stock has outperformed the S&P 500 (^GSPC) by five percentage points over the past month, according to data from Yahoo Finance AlphaSpace, suffice it to say that expectations on the street are skyrocketing for earnings. And that hasn’t gone too well in recent quarters. Therefore, it will not be easy to re-rate the token king’s stock, where an even better valuation could be unlocked due to increased investor interest. However, it can be done, HSBC analyst Frank Lee wrote in a note ahead of the results. “Nvidia’s next big rerating will be driven by a new narrative, as earnings and product roadmap have become less meaningful narratives for the rerating,” Lee said. Lee argued that this new narrative could be the company’s “positioning as the world’s largest contributor to open source AI.” “According to Nvidia, the open source model suite now represents the second most popular category by token generation,” Lee said. “This is critical as open source Small Language Models (SLM) are emerging as the preferred engine for agent AI and on-device applications.” “A push on small language models presents significant gains by lowering the barrier to entry for enterprise inference, expanding the total addressable market (TAM) for Nvidia infrastructure beyond a few frontier labs to millions of individual developers and sovereign nations,” he added.
