NVDA: what now?

I had doubled my NVDA position recently. Sold puts. Bought leaps. I agree that the cryptocurrency related sales did the opposite of masking growth. It made us believe that it was growing faster than it is. The hen has now come home to roost. So the growth rate is not what we thought and now we must recalibrate to that.

I like NVDA long term. There are too many verticals that it is dominating and will dominate, and, yes, these verticals are small but they will one day be big. Not in 2019 but in 2021 or beyond. I think that gaming will reaccelerate and the number of gamers continues to grow. So long term NVDA is a winner IMO.

So what to do. Looking long term, I like it. Looking at the multiple, I like it. The buybacks at these prices will be nice. However, I ask myself will the stock go up faster than my alternatives in the next few months to 6 months? IMO, there isn’t a catalyst until Feb 14 when they report earnings again. Now that management knows what has happened, they will not miss again. They will exceed guidance but will that be good enough? And what will the guidance be on 2/14? What is the next catalyst after 2/14. GTC is in March and there will be new announcements. On a side note: what does the realization of a glut on NVDA’s GPUs mean for AMD’s GPU business? This cannot be good for AMD when it must compete against NVDA’s n-1 generation chip. Also, will we be looking at any inventory write down on NVDA’s large inventory?

So what do I do? with my stock? with my leaps? with my short puts? I do need capital losses to offset the realized gains that I have so far in 2019. I will take the loss on the short puts and switch to another horse for my short put strategy, a stock or stocks that has/have near term catalysts. My shares (if I only sell some) I can’t sell until the end of the month or I would not be able to take the loss (wash sale). I will not sell all my shares but will sell my high cost basis shares to reduce my 2018 tax bill. I will reduce my allocation, and I will keep some shares. I think I will keep the leaps (they are 2021 calls).

Did I make a mistake? Well, the growth rate of the true underlying business (i.e. without crypto) was lower than we thought due to crypto. Everyone said and I said that crypto was not important for NVDA’s long term growth. This was true and still is true. But crypto made growth seem higher and it caused NVDA management (who also didn’t see how much of an impact crypto was having on GPU sales) to overestimate demand and build too much inventory. So now there’s a glut of a previous generation product. So I didn’t see it and Jensen didn’t see and Colette didn’t see it. Now it’s clear. We now also have 2 additional risks with NVDA (the company) that I had not previously considered:

  1. the product cycles are fast due to their fast innovation and product development; this speed of innovation is preventing them to get disrupted but it is a double edged sword because they are disrupting themselves and this can cause inventory to be overbuilt leading to some real business problems.

  2. the lack of visibility into the level of inventory remaining in their channels. There are two of more layers which makes it more difficult to manage their inventory as they introduce a new generation of products. In this case of the crypto there is also a secondary market which exacerbates the inventory problem because when the crypto miners no longer need their GPUs because crypto prices tanked they will attempt to sell their used GPUs adding to the supply.

We are now seeing these 2 risks harming NVDA for 1-3 quarters. In considering these risks, I may need to think again about how much I am willing to allocate to NVDA.

Chris

39 Likes

Chris,

In the end I don’t think it is risky per se to continue holding Nvidia. Nvidia will be back, and it will see better days. It is really opportunity costs to consider. Is Nvidia your fastest horse anymore? Not that you always need a fast horse. NTAP, Cisco, ABMD (a very fast marathon runner, but not a sprinter) have all produced excellent returns of late with much slower growth and without the SaaS type business models.

From these levels, although things may go lower, higher, plus or minus, it is likely that Nvidia will be 20, 30, 40% higher over the next year. I mean this quarter was just the worse. I cannot recall much worse of a quarter for any quality company I have ever held. Things will only get better in 2 or 3 quarters and the market will turn.

But expectations were so, so, so, low for Nvidia that we were thinking 50% to a double as momentum turned Nvidia’s way again after it showed the world that it was not just a bigger and better AMD. Nvidia showed this quarter that not even AI, that all present new technological marvel, is immunity from having a really bad quarter. Data center had no excuse for missing by $29 million, other than business really slowed unexpectedly. So be it.

To me the decision is opportunity cost, not really great risk by continuing to hold (unless your in Leaps of course and margin). I don’t like the lost opportunity costs, so although it was a good thought that Nvidia would beat such low expectations (as so many of us were confident of) it just did not happen and a down market exposed a weakness for Nvidia. Crypto-currency was much more material than thought. Meanwhile, despite a weak market exposing weaknesses, TTD just became stronger.

So do I want to keep money in Nvidia or perhaps take the lumps, and move it into TTD, as an example. I think that example is illustrative. But each investor makes their own choice. In the end, just as Nvidia surprised downward big time this Q, perhaps it will outperform TTD, as an example, over the next year.

Tinker

5 Likes

In the end I don’t think it is risky per se to continue holding Nvidia. Nvidia will be back, and it will see better days. It is really opportunity costs to consider. Is Nvidia your fastest horse anymore?

Yes, Tinker, that’s exactly it. The opportunity cost. In the short- to mid-term, what is likely to move NVDA up more than a general market uptrend? I can invest in the faster horse. Before today, I thought NVDA was that fast horse but it just spawned its ankle. The other factor for me is that opportunity to book a loss (for tax purposes). I now think that NVDA will be a slower horse. And my expectations for the growth rate have been dampened. I will not sell all but I will substantially reduce.

Chris

2 Likes

I had a 7.1% allocation in NVDA. It was once one of my reliable top-3 performers (in a vs. S&P metric, as calculated by my Motley Fool Scoreboard). Lately it’s been in the bottom-3, and after last night it’s going to be solidly last.

I am looking at NVDA for long term trends. And I’ve been trying not to be a short-term focused investor. But I’m still puzzled how they could have missed this badly and not sure what I’m going to be doing next about my holding. I’m waiting for the Fool’s take on this, both the quarterly report summary and the “10% Promise” article they do in these situations.

Yes, Tinker, that’s exactly it. The opportunity cost. In the short- to mid-term, what is likely to move NVDA up more than a general market uptrend? I can invest in the faster horse. Before today, I thought NVDA was that fast horse but it just spawned its ankle. The other factor for me is that opportunity to book a loss (for tax purposes). I now think that NVDA will be a slower horse. And my expectations for the growth rate have been dampened. I will not sell all but I will substantially reduce.

Chri

Yes, for me it may come down to tax strategy, selling for a tax loss. I don’t see any near term catalyst. But the longer term picture is still intact and I would probably buy back once it’s long enough to avoid the wash sale rule.

There’s too much blame on crypto, and not enough on other issues here.

‘Non-GAAP GM decreased by 250 bps to 61.0%, below consensus at 62.8%, mostly on a negative impact from $57M in charges related to sharp decline in crypto mining demand.’

This $57m charge and slowdown in crypto is a minor minor part of their $700m miss in next Q revenue estimates. I think you’re missing the forest for the trees. That’s not why they’re going from up 21% sales YoY to down 15% YoY.

If you think they can grow at 25% going forward the next few years, down from 30%, it’s probably a buy down here.

Another risk is the adoption of deep learning software by very large customers, the better the software gets, the fewer chips you need [all other things being equal]. So, fewer GPUs for data centers in that case. Worth considering especially on this software-friendly board.

6 Likes

Another risk is the adoption of deep learning software by very large customers, the better the software gets, the fewer chips you need [all other things being equal]. So, fewer GPUs for data centers in that case. Worth considering especially on this software-friendly board.

That is a good point, and somewhat goes along with one of Huang’s favorite things to say “the more you buy, the more you save”. At some point, will the customers have all they need on the AI hardware side?

volfan84
still long NVDA, but sharing the thought that NVDA management was caught off-guard by just how much of an impact the crypto bubble had on their business

1 Like

https://mashable.com/2016/01/08/naughty-america-vr-porn-expe…

There is always more.

Always.

Not selling right away if ever.

Cheers
Qazulight

1 Like

“the better the software gets, the fewer chips you need”

My experience has generally been “The newer the software, the better the hardware you need.”

The main risk is for GPU’s to become more of a commodity, so that Nvidia has to make better hardware just to survive, rather than maintain its competitive advantage.

This is not much of a worry yet. Even in a mature area like video games, they just came out with ray tracing, which is now a new competitive advantage, and games will start showing up that use it.

At some point, will the customers have all they need on the AI hardware side?

See point number 1 below:

https://www.pcworld.com/article/155984/worst_tech_prediction…

1 Like