(Source: Telematics Wire)
Nvidia (NASDAQ:NVDA) has shown uncharacteristic growth in the past year, bringing the stock price to its pre-2008 high:
Much of this growth has likely been fueled by earnings surprises:
I recognize Q2 and Q3 gaps as breakaway gaps. Technicals aside, Nvidia has remained impressive by beating estimates three quarters this year, with the lowest surprise being 10% and the highest being nearly 150%.
Compare this level of surprise to those of its industry peers, such as those of Intel (NASDAQ:INTC), which only occasionally make it into the double digits. With breakaway gaps naturally continuing in the direction away from the gap, the recognition of and prediction of these gaps can help investors pinpoint entry times. The natural question, therefore, is whether NVDA shows a pattern in beating earnings expectations.
Before you assume that such a question is purely a technical matter, understand where these gaps come from. The investing community arguably puts too much weight on analysts' opinions. This often leads to price inflations or stagnation.
As a reader of Seeking Alpha, you're probably smart enough to ignore analysts' statements, instead preferring to take a deeper look at the company itself. Indeed, you likely have watched analysts give a buy rating to a stock that has fallen from $50 to $20. Then, once that stock hits the penny stock level, the analyst suddenly says the stock is a sell.
The question is: Where were the analysts saying "sell" at $20? Much like the police, analysts are only helpful after a problematic event. Hindsight is 20/20 for you and I but is a method of maintaining a career for an analyst; it's easy to be correct when you're looking backward.
Still, the opinions of analysts drive markets. Just as an analyst upgrade can influence upward momentum, underestimation of earnings results can suppress a stock. I believe this is what we are seeing with NVDA.
Specifically, analysts underestimate Nvidia. Take January's fall in NVDA for example. We see the exact scenario I outlined above:
As NVDA was falling, Barclays downgraded the stock, stating that NVDA was overvalued. It's easy to say a stock is overvalued when it is continuing to fall. But whether such a call was based on an analysis of Nvidia as a company or pure price action is unclear.
Analysts base their careers on being correct. When a stock is falling, it's easy to say that the stock will continue to fall. But had that Barclay's analyst looked into Nvidia's business trends, he would have seen a clear trend of gobbling up market share in the gaming market, which itself is a gobbling up market share in the entertainment market:
Nvidia is a growing company within a growing industry. It would not be wise to bet against them as they continue to succeed within their marketplace. This is especially true when you consider Nvidia's other major growth area.
I'm speaking of the automotive industry, an industry that Nvidia is fighting tooth and nail to dominate. Although their market share is nothing like that of their gaming market share, Nvidia has grown its presence in the auto industry by nearly double. In the last year, the number of cars equipped with Nvidia tech has increased from 4.7M to 8M.
Revenue from this venture has nearly doubled over the past year:
(Source: WCCF Tech)
At the current rate, which fits a linear trend according to linear analysis, Nvidia can expect to see an extra $8M of revenue from this sector each quarter. Thus, their automotive projects will overtake their datacenter in revenue by next quart. However, the assumption that growth will continue in a linear fashion might be incorrect.
If anything, growth in this industry should be faster, perhaps exponential, as a result of the pipeline of cars set to employ Nvidia's technology. That current pipeline is roughly 30M cars, which is 9 times the market share increase in this year. With $283M being made on the 8M cars, this equates to an expected revenue of $1.1B for the pipeline.
And that's a minimum, as it does not include the autonomous driving sector that Nvidia is attempting to enter. Overall, this equates to a faster-than-linear growth model, though the precise function is unclear at this point. Still, a conservative estimate shows their automotive revenue outpacing their Pro Visualization in 3 years, which should be encouraging to NVDA holders.
I would compare NVDA's entry to the automotive industry much like the entry of Microsoft (NASDAQ:MSFT) into the cloud industry. "Microsoft" does not evoke the word "cloud" when you hear it; nor does "Nvidia" evoke "automotive"… yet. Just as "Apple" is now associated with both PCs and smart phones, Nvidia should soon be associated with both "gaming" and "cars." (And Microsoft will soon be associated with both "Windows" and "Cloud.")
Nvidia has a fight it must win before gaining that association. But much like Microsoft, Nvidia is a veteran of many wars, and early feedback has shown Nvidia to be a successful player in this game. With Nvidia having already won the gaming tech war, Nvidia can now internalize the fact that its competitors are not AMD and Intel so much as the companies trying to prevent Nvidia from gobbling up automotive tech market share.
And the best way for Nvidia to do so is by improving its in-car tech. With a gaming mindset, Nvidia understands users' needs as concerns interactivity. Nividia has proven this in its computing and rendering for the displays in Tesla (NASDAQ:TSLA) cars.
One issue that Nvidia is better set to handle - due to its gaming background - is the display of maps. To quickly redraw a map after being moved or zoomed by the user, the processor must access the data and then render it as a 3D image. Clearly, this is no different than gaming, for which speed is a priority.
Nividia will likely continue to innovate in this regard, while competitors simply up processing power. For example, future displays will likely be switching from raster maps to vector maps. In contrast to the maps rendered as images, such as those you see via Google Maps, vector maps reduce superfluous details, thereby cutting down on processing and rendering time.
Nividia is no stranger to vector maps and its processing. Although I had trouble locating a review that specifically mentioned vector maps, this review from 2014 tests Nvidia's Quadro K6000 against competitors on tasks including vector maps. The review, unsurprisingly found the Nvidia GPU to beat its closest rival at the time.
Indeed, investors have reason to be optimistic of Nvidia's future in the automotive industry. But for now, the main revenue source is still gaming. And - bringing it back to the main topic - the earnings surprises are likely due to the growth in the gaming industry.
While the gaming revenue continues to sit atop Nvidia's throne, we should attribute any patterns in earnings reports or earnings surprises to the growth of the gaming industry and Nvidia's growing market share in this industry. Thus, the study that follows holds as long as those two factors are still at play in the future of Nvidia. We are about to investigate the earnings patterns and resulting stock movements in Nvidia.
The earnings surprise and subsequent movement of Nvidia prompted a look into the earnings patterns of Nvidia. A quick overview of the earnings shows that Nvidia might outperform in Q3 and Q4 and underperform in Q1 and Q2:
I ran the data - including all data from Nvidia's IPO to now, constituting 17 years of data. I found the following results:
Clearly, these are strong results. On average, we would expect the quarterly earning periods to perform equally as well as the stock at any other times. If earnings were to bring only volatility and not consistent surprises in one direction or the other, the volatility would still not apply to overweighting an outperform or underperform rating.
In short, we see a consistent pattern of NVDA outperforming during Q3 and Q4 earnings periods and underperforming during Q1 and Q2 earnings. Because Q1 has the strongest negative reaction, one trading strategy is to remove Q1 from our holding period. The result is clearly in our favor (NVDA_Seasonal represents our strategy; NVDA_HOLD represents buy-and-hold of NVDA; SP_HOLD represents buy-and-hold of the SPY):
The strategy of selling into May and buying back on June 1 more than triples our return. Unlike other seasonal strategies in which we must drop the stock, this one brings in so much extra profits that it makes up for any extra taxes you must pay for not holding more than a year. The above strategy's return includes dividends.
But can we make this strategy better? If we are going to drop the stock anyway, we might as well avoid the Q2 earnings (that of August) risk. This gives a new strategy of holding NVDA from September to May:
Now we have quintupled our return on NVDA. The above is dividend adjusted. But can we make it better?
Yes, because the above strategy throws some of the baby out with the bath water. The above strategy outperformed the first only because it avoided Q2 earnings. Thus, the ideal strategy seems to be dropping NVDA in both May and August, holding at all other periods:
And the result is our best earnings-based holding strategy, with an amazing 850% improvement over buy-and-hold.
With Nvidia looking better with time, dropping NVDA from your portfolio for any time becomes a difficult decision. But I have evidence that NVDA does indeed underperform two times per year. And as stated previously, I believe this is unrelated to NVDA's financials and more related to analyst estimates influencing pre-earnings sentiment.
My last analysis on NVDA was correct, to many NVDA fans' chagrin. I recommended shorting or selling at a time when NVDA showed no signs of slowing down. The stock dropped immediately:
I've also run similar studies on other stocks, accurately predicting significant turnarounds such as predicting Apple's fall in November, predicting GoPro's fall, and predicting Disney's October surge. These are not examples of skill; they are simply predictions based on statistical findings and mathematical valuations. For NVDA, I am recommending the strategy above for the same reasons: Unlike analysts, data do not lie.
Overall, NVDA is a stock with great growth potential. Nevertheless, taking two months' vacation once per year can allow you to avoid NVDA's inevitable dips. Before you say, "You cannot time the markets," please do yourself the favor of reviewing the above returns and asking whether the potential 850% gain is worth the risk.
At the very least, you can set your dollar cost averaging strategies to ensure entry at the probable low points: June 1 and September 1. Or, if you want to test the waters, buy call options in addition to holding the stock for these times. Another possible strategy is buying put options on May 1 and October 1.
I do believe that the market environment make NVDA a good investment. But I do not believe that investment is equivalent to "buy and hold." You can invest smartly with big data.
Request a Statistical Study
If you would like for me to run a statistical study on a specific aspect of a specific stock, commodity, or market, just request so in the comments section below. Alternatively, send me a message or email.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.