2023-11-29 19:16:11 ET
Summary
- PRST has dropped significantly after an initial hype phase due to a couple of unimpressive quarters.
- PRST is backed heavily by insiders and other large investors who want to see common shares succeed.
- Due to the heavy investor backing, I don't believe that PRST is at great risk of crippling levels of dilution seen with other small cap de-SPACs in similar financial shape.
- The company is hinting at heavy levels of growth as the AI requires less human intervention, and it's deployed at more locations. However, revenue forecasts are lacking.
- I have a price target of $2.00, representing a 200%-300% upside based on current prices. I expect upwards revisions to come upon the company achieving operational milestones and with improved market sentiment for small caps.
Artificial intelligence has been one of hyped up sectors of 2023 thanks to the rise in popularity of ChatGPT. Both the positive impacts and detriments of AI are becoming more apparent in popular culture. For a while early in the year, Presto Automation Inc. ( PRST ) was caught up in the positive side of the technology as it ran on sector and company-specific hype. However, after a couple of disappointing quarters along with a weak market for small cap tech and tax loss selling season, the stock has cratered to barely one-tenth of the price it was just five months ago.
It's gotten to a point where I feel comfortable calling it my top AI pick for 2024, even with some financial challenges that must be overcome. PRST merely returning to a market cap seen earlier in the year would require a several hundred percent return. I am submitting this article as part of Seeking Alpha's 2024 Top Picks competition . I figure to be a top pick, you have to go big or go home with a high beta stock that has as much of a chance to be a 10-bagger as it does a dud that eventually erodes to zero.
I believe that part of the reason why PRST has taken such a dive is that it's getting lumped in with numerous other failed de-SPACs and recent IPOs that have turned gangrene from too many rounds of toxic dilution. Micromobility.com Inc. ( MCOM ), formerly called Helbiz, being one such example which I warned investors about several months ago. Having written about and studied numerous small caps with failing business models and exploding share counts, I can understand why investors might fear that PRST will fall into this trap as it does have some similarities to MCOM. However, it will be the differences from these types of indefinitely unprofitable ventures that make PRST such a compelling buy. I have had a fair amount of success in being able to separate the wheat from the chaff when it comes to small cap stocks. I believe that PRST has a good chance at being wheat while currently being valued as an empty corn husk.
I initially became interested in the Presto not because of the stock, but because of its cheap warrants ( PRSTW ), which were trading at less than $0.10 when the stock was in the $3-$4 range. I have a long history of writing about de-SPAC warrant opportunities on Seeking Alpha and this one fit my investment strategy well. Despite having recently sold most of my warrants at more than a 50% loss, this strategy did end up working somewhat as I used the proceeds from the sale to buy the stock after an 80% decline. By the time the warrants become in-the-money, the stock will have had to move greater than 1,000%. PRST offers a superior risk-to-reward trade-off to PRSTW right now. The warrants become a potential buy again at the right price after the stock has significantly bounced from these current levels.
Presto Voice: Compelling "good type" of AI with a use case that could save the restaurant industry millions
Presto has been servicing the needs of restaurants for more than a decade with Presto Touch. This platform functions as an all-in-one server handheld or tabletop guest ordering, payment, customer personalization and gaming device. But that's not what makes PRST interesting as a speculative investment. The company's future prospects rely heavily on the success of its new Voice AI platform for drive-thrus. You can view Touch as the precursor that allowed Presto to learn about the restaurant industry. Challenges such as POS and menu integration and the complex dynamic between brands and their franchisees. Presto therefore knows more about the restaurant industry and its specific needs over the average AI developer.
The AI industry is controversial in that it offers plenty of pitfalls along with benefits for society. People worry about deepfakes causing chaos in politics and society at large. Or about job loss as robots take over a greater portion of good-paying working class jobs. However, I believe that the automation of drive-thrus offers only upside. Employment in fast food is almost always low-paid, unskilled labor that is performed by young, inexperienced and/or lowly motivated employees. I've never worked in such an environment. However, I can imagine that the drive-thru portion of it - with people screaming orders non-stop at you for hours - being the worst part of the job. Drive-thru AI is set to reduce employee errors and improve morale. There is a cost savings component to it like most AI infiltrating the workforce, but what makes Presto Voice unique is the ability to increase revenue through it having a better ability to upsell than demoralized, tired and inexperienced workers.
This has been shown in initial tests of drive-thru AI technology. There is a reported $35,000 per location cost savings, a number that will only increase with time as government-mandated minimum wages go up. That's coupled with a reported 6% increase on average in revenue thanks to better upsell ability by AI over tired, minimum wage zombie workers. Pretty much all of the biggest fast food restaurant chains are trying some form of AI technology in the drive-thru. Some restaurant chains are trying to develop their own technology, while others are using third party providers such as Presto. While these internal options only drive 80% order accuracy, Presto claims a 95% order accuracy rate. However, this does come with some caveats. On the company's latest investor call it confirmed that only 30% of orders are currently autonomous. The remaining 70% require some human intervention on behalf of the company. Presto uses human-in-the-loop to train the models while ensuring orders remain accurate at the client-facing level.
The automaton of the drive-thru is unambiguously positive for the restaurant industry and the technology exists to make it happen. However, the actual implementation and perfection of the technology will require some kinks to be worked out. AI, by its very nature, requires a learning curve. Such as the one requiring human-in-the-loop mentioned in the above paragraph. This learning curve, education of the restaurant industry and onboarding process will result in some financial challenges for companies like Presto in the short term.
Financial and competitive challenges exist, but upside outweighs downside risk at this valuation
One positive aspect of PRST that has kept me bullish is Chairman, founder and former CEO Krishna K. Gupta being active, engaging and helpful with shareholders and other members of the investment community on Twitter , among other places. Unlike other management teams who engage with retail shareholders for the primary purpose of trying to promote the stock, I believe that he is attempting to help people to learn about the business. However, he also gets caught up in talking about the "naysayers". PRST's stock price would not be in the position it's in without some serious financial challenges that it must overcome. However, I believe the company is at the trough, both in terms of stock price and operating performance. Making this the ideal time to invest.
Immediately following the release of the Q1 results ended September 30, 2023, the stock tanked 30%. Reading the financials, one can understand why. Revenue for Q1 2023 was $4.9 million, in line with the company's forecast but down from $7.8 million in the previous year. Gross margin was almost non-existent at $203,000. While the company managed to pull an accounting profit thanks to the change in fair value of its derivatives, the operating loss was $13.2 million. Most disappointing to me is that in all of this purported traction that the company is seeing, guidance for Q2 remains the same as it did for Q1 - $4.8 million to $5.0 million in revenue.
However, the appearance of poor operations is due to the stagnating legacy business. Revenue is down and margins are poor as the company figures out its future for Presto Touch and its next generation, Flex. The company also enacted cost-cutting measures which it expects to save $3.6 million per quarter by the middle of 2024. The stagnant revenue expected in fiscal Q2 would likely tell a clearer story if it was broken down by Voice versus Touch. Though for whatever reason, the company is being vague with revenue expectations on Voice while being very open about its forecast on the rapidly increasing number of locations where Voice is deployed (more on that later).
The balance sheet doesn't do much to help calm investor fears. The company ended September with $18 million in current assets against $63 million in current liabilities for a $45 million working capital deficit and negative shareholder's equity to the tune of $36 million. The working capital deficit is due to a $55 million loan from Metropolitan Partners that is classified as short term debt. Metropolitan has been amicable to amending certain covenants of the loan. The company actually improved its balance sheet from the previous quarter thanks to Gupta and Remus Capital-affiliated syndicate of investors making a $7 million investment in the form of common stock. That, along with Metropolitan Partners being a supportive lender up until now, leads me to believe that the company has sufficient backing of large creditors and shareholders to make it to cash flow positivity without drowning existing shareholders in too much dilution.
PRST has some competitive challenges along with financial ones. While third party AI solutions are performing better than internal technology created by restaurant chains, Presto isn't the only third party provider out there. PRST is seen as a junior or lesser version to SoundHound AI, Inc. ( SOUN ), which also seeks to disrupt and improve the restaurant industry with AI-based solutions among its suite of AI tools for various industries. Given that Presto comes from the restaurant industry while SoundHound is trying to expand into it, I feel that the perception of Presto being the inferior competitor specifically for the restaurant industry is misplaced. Though a comparison of the numbers tells you why some investors might think that way:
Data generated from SEC filings and my own calculations
While SOUN has a similar cash burn to PRST, its balance sheet is in much better shape and has been experiencing revenue growth on nearly three times the base. That being said, SOUN also has 14 times the market cap and 7 times the enterprise value. That leads to PRST having a revenue multiple of 4, 2.5 times lower than SOUN's 10 and one of the lowest multiples you'll find out of the legitimate companies in the AI sector. If PRST can demonstrate a turnaround in its financials, it can considerably improve its valuation as revenue growth would then be multiplied by a multiple of 10 or more. It can also counteract the narrative that it is inferior to SoundHound by showing itself to be a leader in restaurant industry-focused AI solutions. That may further increase its valuation multiples.
Presto is not a dilution scam even though it's being treated as one
A dilution scam is a term for a company that undertakes numerous toxic financing deals in support of a business that is really more of a science project than a business or to pay management salaries. Instead of operating with the intent to earn a return for shareholders, these type of companies view their shares as currency and their shareholders as suckers. Not all of these types of companies start out that way and may have been operating under the assumption that one day they will be profitable. But they unfortunately fall under the weight of poor operating performance until the only way they can keep their business afloat is through numerous toxic financings.
In the previous section, I outlined the significant challenges PRST has with stagnant revenue, poor margins, high cash burn and negative working capital. I reiterate my analysis on MCOM, which at the time of writing, had a negative working capital similar to PRST today. Along with stagnant revenues, poor margins and heavy cash burn. PRST will undoubtedly need to raise funds in the near term in order to finance the business until it is profitable. Just like MCOM has done. The difference is how each company is going about raising those funds. Here is a comparison of PRST to MCOM's growth in shares outstanding over the past year to hammer home the point that I am trying to make:
MCOM's shares outstanding has exploded a mind-numbing 100 times from less than 3 million (adjusted for a reverse split) to over 300 million in 2023. Meanwhile, PRST's share count has grown a comparatively modest 17 million from 50 million to 67 million over the last year. Instead of toxic financing from questionable small cap-focused institutional investors at a discount to the market, PRST is getting financed - often by insiders and other large shareholders - at a premium to the market. Gupta, Remus Capital and the rest of the investors who have been financing Presto to date have every incentive to see the company thrive and shareholder value to increase. The "investors" in MCOM don't care about the company's fortunes one way or another. As long as there is enough liquidity to sell the shares they receive at ever-lower prices, they will continue to fund it.
PRST investors will move boulders up mountains before they let the company undertake toxic financing similar to MCOM because that will wipe out the significant amount of money they put into the company alongside retail shareholders. Certain PRST investors - including Gupta - have extended the lock-up period of their shares to December 23, 2024. 31.8 million shares of PRST are subject to this extended lock-up. That's 47% of the total share count, but it was 64% prior to any dilution. Just to demonstrate the impact of the relatively tame dilution has had on these shares so far.
Based on this, I have a high degree of confidence that PRST will strive to keep its share count at a maximum of 100 million (excluding the impact of currently out-of-the-money derivatives). That leaves 33 million shares to be issued. There are 180 million shares authorized, so that is the absolute ceiling that PRST can issue (including shares from the exercise of derivatives) before requiring to request an increase at a future AGM. The company recently filed a $75 million mixed shelf offering. My guess is that this is the amount the company estimates that it will need to raise until it has a reasonable chance at attaining positive cash flow.
PRST also differentiates itself from a company like MCOM because it has a business that can profitably scale without a lot of capex while MCOM does not. MCOM's last mile e-scooter business requires fleet for any new market it wishes to enter. Fleet that is subject to depreciation, repair and replacement. PRST requires system integration, testing and a learning curve that one would expect for any SaaS type of business model. But once that happens, it's high margin from there. While I expect MCOM to continue to struggle to find its way towards profitable growth, that path should be achievable for PRST.
Figuring out Presto's revenue and path to profitability
Along with the financials, PRST issued an operational update on November 21 that included some important information on Presto Voice:
As I mentioned previously, currently 70% of orders taken through Presto Voice require human intervention. That's understandable given the nature of AI. The system has to learn how to take orders at each individual location. Burger places would have their own nuances compared to chicken places, for instance. Both myself and the company expect that to improve with time, which will increase site margin. Right now PRST is just below the breakeven point at each site, and expects positive margin by June of next year when orders not needing human intervention achieve a 50% rate. Therefore I expect margins to drastically improve from PRST's fiscal Q1 moving forward.
PRST also disclosed its forecasted number of locations in the eight quarters for the period January 1, 2024 to December 31, 2025. While the company is currently in over 400 locations, it expects to be in 562 locations in calendar Q1 of 2024 with that number growing fairly steadily every quarter until it reaches 2,610 locations by the end of 2025. The bulk of this location growth will be from restaurant brands that already have master service agreements or pilot programs for which Presto Voice has already been installed. So as long as the company doesn't lose any more clients like it did with Del Taco, these numbers should be very achievable.
The last component needed to come up with a reasonable revenue estimate would be the expected revenue per location. The chart below is from the company's April 2023 investor presentation:
Based on 15,500 locations, the company estimates that it could achieve $200 million in annual revenue. That works out to approximately $13,000 per location. However, the numbers estimated in the total addressable market and serviceable addressable market lead to $30,000 and $17,000 per location, respectively.
On the earnings call , the CEO mentioned that approachable annual revenue is in excess of $100 million. With this number being defined as total revenue from locations of restaurants groups for which Presto has a master service or plant agreement and where Presto Voice has not yet been deployed. An analyst asked the company that based on 1,200 locations, would that lead to $25 million in ARR? That number would assume over $20,000 in revenue per location. Unfortunately, the CEO did not answer the question, leaving it up to the analyst to come up with their own estimate.
Using 2,610 locations and multiplying that by $13,000 in revenue per location leads to approximately $34 million in an annual revenue run rate. Using $20,000 per location leads to $52 million in an annual revenue run rate. Since I started first researching this stock, I have seen per site revenue assumptions that generally fit between the $10,000 and $20,000 mark. So I will use $15,000.
$15,000 in revenue per location leads to $40 million in an annual revenue run rate based on 2,610 locations. Multiply that by a 10x revenue multiple and divide by 100 million shares I assumed from above and that comes out to a $4.00 target. Easy right? Well, it's not that simple. The final step is to assess what I feel are the biggest risks associated with PRST.
Assessment of risks and final comments: price target of $2.00
I have identified the four biggest risks I see with investing in PRST. Those four are: ambiguity around future revenue growth, financial risks related to the negative working capital and overleveraged balance sheet, technological risks and competitive risks.
Revenue : I believe that some of the drop in PRST's price is self-inflicted rather than originating purely from outside negative forces. The company's inability or unwillingness to provide some kind of guidance on revenue beyond the next quarter when it has such strong expansion plans is creating unnecessary ambiguity. It doesn't help that it doesn't split out legacy revenues from Presto Voice alongside claims of Voice being deployed in more and more locations seemingly daily. The market only sees that overall revenue is down year-over-year and forecasted flat going into the next quarter. Perhaps due to the high human intervention rate seen in the early stages of deployment, revenue is far too volatile to reliably predict right now and management doesn't want to be too far off with guidance. As an outside shareholder and market participant, my response to that would be anything is better than $4.8 million to $5 million in revenue for the next quarter, and that's it. Even if a forecast is in an excessively wide range for revenue in calendar 2024, I would like to see it. There is only so much credence people will give to off-hand comments about ARR during an investor call.
Balance sheet : The elephant in the room here is Metropolitan Partners and its $55 million loan. To date, Metropolitan has been a friendly and patient lender. Perhaps it sees profitable growth on the horizon that is driven by imminent site margin positivity and location growth. However, if PRST was unable to perform or Metropolitan in some other way became less friendly, it could lead to insolvency or heavy dilution for PRST shareholders. I reiterate that authorized shares are 180 million and company insiders are heavily incentivized to minimize dilution. So dilution risk is limited, but bankruptcy risk is not. The offset to this risk would be asking what exactly would Metropolitan's end game be if it were to force PRST into bankruptcy? There are not many tangible assets to speak of. If it took over the business, then it would need to manage and fund it by itself. The easiest path for Metropolitan is to let PRST management do what it needs to do in order to become cash flow positive. So I am reasonably confident that Metropolitan will remain friendly or at least patient for the foreseeable future.
Technology : I am not too worried about the AI behind Presto Voice. The AI being independently functional 30% of the time at this early stage seems reasonable to me given that Presto claims and that most restaurants require at least a 95% order accuracy. This is not a situation where people can troll ChatGPT then post the answers on social media for fun. A drive-thru order is either right or wrong. If the technology was complete garbage, it wouldn't be able to attain Pete Rose's batting average in the first place. Therefore, I remain quite confident in the company's goal to see half the orders free from human intervention within the next several months.
Competitive : This is another issue that exists, but I don't believe it to be a major risk. While people debate the superiority of SoundHound to Presto, keep in mind that between a handful of small cap players and some internally produced options, the AI integration of drive-thrus is nowhere near a saturation point. At least in the short term, it doesn't matter if Presto might be superior or inferior to the competition. Each competitor is going to carve out its own market. For instance, based on ARR estimates, PRST may not even need to expand much beyond its existing client base for Presto Voice to become a viable and profitable investment. It just needs to deploy the solution in all locations. $100 million in revenue leads to a billion dollar company at a 10x revenue multiple. As the industry and AI products mature, I do expect saturation. After a few years, it will be impossible for new players to emerge unless they really undercut existing contracts. The AI will become commoditized but at the same time personalized. Presto and all the other players will have AI that's very, very good at what it does. A restaurant chain that has used a certain company's AI for years that runs smoothly will be reluctant to switch over to a new one and start the learning curve all over again unless given very clear incentives to do so. Any AI that is not strong enough to hold high accuracy after certain short period of time will fail. But that's due to an inability to perform, not due to competition taking over. After several years, I expect consolidation in the industry, but not for individual companies to die away from competition once they have already acquired and pleased their clients.
Taking all of that into consideration, I feel a 50% discount on the $4.00 estimate to come to a $2.00 target is fair. Based on 100 million shares outstanding (currently 67 million), this is a $200 million market cap. This market cap is lower than the market cap seen at the peak of PRST's hype during the summer. So PRST can easily trade higher than $2.00 purely on market hype alone. But given the stock price is in the $0.50-$0.60 range right now, a $2.00 target already represents a 200%-300% upside. I feel comfortable with that target for my top AI pick in 2024 and can revisit the story if and when it gets to that price. One can assume if it does reach $2.00, some of the risks around operating and financial performance will have been alleviated and it's worth reevaluating for an increased target.
Editor’s Note: This article was submitted as part of Seeking Alpha’s Top 2024 Long/Short Pick investment competition , which runs through December 31. With cash prizes, this competition -- open to all contributors -- is one you don’t want to miss. If you are interested in becoming a contributor and taking part in the competition, click here to find out more and submit your article today!
For further details see:
Presto Automation: My Top AI Pick For 2024