MRAM - Everspin Technologies: Hanging On A Cliff

2024-01-16 04:56:06 ET

Summary

• Everspin Technologies stock fell after being unable to overcome resistance, but support helped stem the slide, although it may still be in danger.
• MRAM beat estimates in its latest report, but the stock still sold off because of signs demand is getting softer.
• The Company has benefited from rising interest in AI and the stock has risen accordingly, but the exact role it will play has yet to be determined.
• There are a number of reasons why it may be time to take some chips off the table in order to reduce exposure, including a possible breakdown.

Everspin Technologies ( MRAM ) was one of the stocks to benefit from rising interest in artificial intelligence or AI. The stock rallied for months, starting in May when AI took it to another level, but that rally has come to an end, in part due to a selloff in the stock following the most recent earnings report. The stock has gone sideways for over two months and it may be priming itself for another move lower. Why will be covered next?

Why MRAM is hanging on a cliff

The stock had rallied higher for months at the time, but a previous article from mid-October rated MRAM a hold after concluding that the rally was unlikely to have much further to go for several reasons, including the possible presence of resistance nearby. This assessment proved to be on point because the stock matched the 52-week high of $10.50 a few days later on October 19, before the stock proceeded to decline as shown in the chart below.

The decline included a drop of 11.4% to as low as $8.33 on November 2 following the Q3 report. However, as shown in the chart above, the stock has not gone any lower than the post-earnings low. The stock has gone up and down, resulting in a stock moving sideways. There may be a good reason why this has happened.

As mentioned in the prior article, there was reason to believe the stock did not much further to go by mid-October. One of the reasons given in the article was that the stock was on a collision course with a potential resistance level. The article identified $10.67 as a price point where resistance could be present because it equaled the 61.8% Fibonacci retracement of an earlier downtrend as explained in greater detail in the prior article.

This turned out to be correct in hindsight with the stock twice topping out at $10.50, which is $0.17 below $10.67, before retreating. Similarly, there is reason to believe the reason why the stock went as low as $8.33 and no more is because support may not be far away. If the recent drop in the stock is a retracement of the move up from the July 2022 low of $4.71 to the October 2023 high of $10.50, then the 38.2% Fibonacci retracement is $8.29.

This is within pennies of the recent low of $8.33, which could explain why the stock has not gone lower in the last few months. It's because of support. Still, support can be broken, especially if the stock keeps challenging support like it has been doing for over two months now. So while support has held, the possibility of a breakdown below support cannot be ruled out with the stock hanging out so close to support.

It's also worth noting how the highs are going lower once we connect them with a trendline. This suggests the bounces off of support are losing strength, which may be an indication a break below support is becoming increasingly likely. At the moment, the stock finds itself between support and resistance, although much closer to the former.

Why the stock dropped off

A move lower becomes more likely if MRAM continues to disappoint as it did in the most recent report. As mentioned earlier, the stock dropped by double digits after the release of the Q3 report. MRAM actually beat estimates for the top and the bottom line, but it fell short elsewhere. For starters, MRAM had guided for GAAP EPS of $0.01-0.06, but MRAM earned $0.11 on revenue of $16.46M, or $0.05 above even the high end of guidance.

Gross margin and thus earnings benefited from an increase in licensing, royalties, patents and other revenue to $2.9M, up from $0.7M a year ago, because licensing comes with higher margins. MRAM finished the quarter with cash and cash equivalents of $34.9M and no debt on the balance sheet. The table below shows the numbers for Q3 FY2023.

Notice how the bottom line declined QoQ in Q3 FY2023 versus Q2 FY2023 even though the top line grew. It's therefore worth mentioning that Q2 FY2023 benefited from a $2M tax credit, which boosted the bottom line. Otherwise, GAAP EPS in Q2 FY2023 would have been more like $0.08, which puts the QoQ comparisons in a different light.

Source: MRAM Form 8-K

However, the bump in licensing revenue was a one-time thing. Guidance therefore calls for Q4 FY2023 revenue of $15.4-16.4M, an increase of 1.3% YoY at the midpoint, and GAAP EPS of $0.01-0.06, which is about flat YoY and down QoQ.

Source: MRAM Form 8-K

Furthermore, there is some evidence demand is getting softer. For instance, the backlog, which stood at over a year until fairly recently, has gone down. From the Q3 earnings call:

I guess the way, I would describe it, so what's happened is the lead times were 52 weeks. They came down to about 30-something weeks, and then we reduce them further to about 27, 26 weeks. So we saw lead times, come down. And then in addition to that, one of the things you'll see industry-wide that you might have noticed, is that the supply constraints loosened up as well,"

A transcript of the Q3 FY2023 earnings call can be found here .

MRAM is projected to end FY2023 with GAAP EPS of $0.39 assuming Q4 EPS comes in at the high end of guidance at $0.06. This would give MRAM a P/E ratio of about 22 since the stock closed at $8.54 on January 12. In comparison, MRAM earned $0.29 in FY2022 and $0.22 in FY2021 and it reported a loss of $0.45 a share in FY2020. This represents an improvement of $0.84 in the last three years.

MRAM has improved earnings every year since it went public in 2016 and when it finished the year with a loss of $3.53 a share. If we then assume earnings grow by an estimated 30% per year on average and TTM EPS of $0.39, then this would imply fair value of $11.70 for MRAM. In comparison, the average price target on Wall Street is $12. This is well above the current stock price.

Will MRAM be a winner in the long run?

As mentioned earlier, the stock has risen, especially after investors started going after stocks that could potentially stand to benefit from AI. MRAM, the company, fits the bill as the leading supplier of magnetoresistive random-access memory or MRAM. MRAM, the technology, is a type of non-volatile random-access memory, which could be of use in certain AI applications, particularly those at the Edge, due to its ability to retain data, which allows for, for instance, continuous monitoring of model parameters without too much power consumption.

The fact that MRAM, the company, has made great strides in terms of earning a profit further strengthens the belief MRAM is indeed benefiting from rising industry adoption of AI. However, it needs mentioning that the perceived benefits from AI are mostly on paper at this time. In addition, MRAM, the technology is not necessarily the only type of non-volatile random-access memory to gain from AI.

MRAM, the technology, has many other competing types of non-volatile random-access memory and new ones continue to appear, each with their specific pros and cons. The race as to who will come out on top, if any, remains to be determined. It's not out of the question that when all is said and done, MRAM, whether the company or the technology, loses out in this competition.

Investor takeaways

I remain neutral on MRAM. MRAM is coming off a rather long rally that saw the stock multiply in value. The rally shifted to higher gear when interest in potential AI plays jumped in May. The stock was arguably due for a correction and this is what ultimately happened in October. The last few months have been mostly corrective in nature.

However, while MRAM has made progress in terms of improving profitability, it remains to be seen whether that will continue in the coming years. Keep in mind MRAM has not been around that long as a listed company and its future performance may be quite different from the last several years. The future of MRAM as a technology is still up in the air and there are no guarantees it will come out on top. MRAM could very well lose out to something better.

The recent signs that suggest demand is getting softer is therefore something worth keeping an eye on. There is still a substantial backlog, which means the headline numbers won't see much of an impact from softening demand for some time because MRAM can work on fulfilling previous orders. But if the backlog continues to shrink due to a lack of new orders, the lack of demand will find its way back to earnings.

The charts also suggest caution is warranted. The stock was unable to get past resistance and in combination with a disappointing earnings report was enough to bring down the stock a notch. While the stock has found support and support has held its ground for a couple of months, the fact that the stock is staying close to support means the stock is not out of the danger zone. The stock is still very much in a position to challenge and breach support.

Longs can breathe easier if the stock moves away and if it increases its distance to support, but not any sooner. At the moment, MRAM finds itself between support and resistance, although much closer to the former. If the stock is unable to breach either, continued sideways action is in store. A catalyst may be needed to break the stalemate if this were to happen. This could be in the form of the next quarterly report, but that is not likely to arrive until early March.

Bottom line, there are several reasons why retaining a neutral stance is warranted. Whether it is the future of MRAM in the long run as a technology, the possibility of an imminent breakdown in the stock or the uncertain state of demand, there are several reasons why bulls should probably not expect to see too much of a return anytime soon. This can change, but only further down the road.

For further details see: