Is AAOI's P/S Ratio of 18.5 Expensive? Model Says 'Undervalued by 13%', Short Sellers Say '67% Bubble' — Who's Lying?

Published on 2026-08-17Last updated on 2026-08-17

Abstract

The crucial question facing investors is whether the demand for its high-speed optical products and its planned capacity expansion can support the current revenue expectations. If this assumption proves untenable, the current valuation might resemble a value trap rather than a sound investment opportunity.

Applied Optoelectronics' stock has delivered many years of very strong upward momentum, but its low overall Value Score suggests the current share price may already reflect much of the market's optimism.

Over the past 5 years, Applied Optoelectronics' returns have been approximately 19 times the initial investment, which further focuses attention on how much upside the company may have left.

Recently, surging demand for the company's high-speed optical products (used to support AI and cloud data centers) may have raised investor expectations. However, any slowdown in demand or hiccups in planned capacity expansion could quickly affect what investors are willing to pay.

Applied Optoelectronics passes only 1 out of 6 valuation checks, so even considering its growth prospects, the company appears overvalued rather than undervalued on the whole.

The question now is whether Applied Optoelectronics' current valuation offers enough runway for reasonable returns after such a strong run-up.

Applied Optoelectronics returned 519.1% last year. Here's a look at how it stacks up across the Communications industry.

Is Applied Optoelectronics a Bargain Now?

For Applied Optoelectronics, the Price-to-Sales (P/S) ratio is a useful metric because the company is focused on revenue growth rather than established profitability.

Applied Optoelectronics' P/S ratio is around 18.5x, significantly higher than the Communications industry average (~2.4x) and higher than the peer average (~7.0x). On a simple comparison, investors are paying much more per $1 of sales for this company than for many other communications stocks.

However, a more justified P/S ratio, considering the company's growth prospects, margins, size, and risk, is around 29.3x, which is still far above the current P/S. Despite the strong recent rally driven by AI-related policy news and revenue updates, the model indicates the current price does not fully reflect market expectations for Applied Optoelectronics based on revenue.

On a P/S basis, Applied Optoelectronics stock appears undervalued relative to where it should typically trade given its growth and risk profile.

The Applied Optoelectronics Narrative: What Must Hold True to Justify Today's Price?

The Simply Wall St Narratives series picks up Applied Optoelectronics' valuation puzzle, outlining what assumptions in growth, margins, and earnings must hold for the stock to be worth significantly more or less than its current price. Each article treats Applied Optoelectronics' implied fair value as a theory about the company's business that can be reassessed over time, not a one-time snapshot.

Currently, there is a stark divergence in views on Applied Optoelectronics, with a vast chasm between bullish and bearish theses on the stock.

Bull Thesis: Undervalued by 13%

"Growing demand from AI/machine learning workloads, video streaming, the Internet of Things, and the ongoing transition from copper to fiber in network infrastructure creates persistent tailwinds for the industry..."

Bear Thesis: Overvalued by 67%

"The case for going short is equally clear: the stock already reflects most of its upside potential..."

Summary

Applied Optoelectronics appears undervalued based on a customized Price-to-Sales (P/S) metric, but its low overall Value Score suggests most other metrics are not giving a clear endorsement. Recent high volatility in the share price implies market expectations are already elevated, warranting extra caution on relying on the P/S metric alone. The key question for investors is whether demand for its high-speed optical products and planned capacity expansion can support the current revenue expectations. If this assumption fails to hold, the current valuation might look more like a value trap than an investment opportunity.

Related Reads

Google and Meta Called Out for Benchmark Gaming

Recently, analysis firm SemiAnalysis accused tech giants Google and Meta of "benchmark gaming" with their AI models Gemini 3.8 Flash and Muse Spark 1.3. The accusation stems from a dramatic performance drop between two versions of the Terminal-Bench evaluation for AI agents. On the older, public Terminal-Bench 2.1, Gemini 3.8 Flash scored 89.4, ranking second and beating GPT-6 Astra, while Muse Spark 1.3 scored 88.8. However, on the newly released, more secure Terminal-Bench 4.0, their scores plummeted to 19.1 and approximately 33.3 respectively, far behind competitors. SemiAnalysis argues this indicates "benchmark contamination," where companies train models not on the public test questions themselves, but on expensive, privately purchased training data specifically designed to mimic the benchmark's style. This has evolved into a lucrative industry, with specialized firms selling tailored training tasks for thousands to hundreds of thousands of dollars. The article specifically points to Datacurve, a company that both sells expert coding data and runs its own benchmark (DeepSWE), where the accused models also performed well. Meta's Chief AI Officer, Alexandr Wang, dismissed the claims as a "silly argument," pointing out similar performance drops for other models like GPT-5.6 Sol. He stated Meta never claimed Muse Spark 1.3 was as powerful as top-tier models, only that it offered better value. The report concludes that this is the inevitable fate of all high-quality public benchmarks—they become "gamed" over time. The proposed solution of private, high-quality benchmarks comes with a significant downside: it would erode public transparency, turning open rankings into marketing tools and leaving developers without a common, fair measure to compare AI models.

marsbit2h ago

Google and Meta Called Out for Benchmark Gaming

marsbit2h ago

Crypto's Nouveau Riche Strikes Gold in the Real World: Coinbase Co-founder's Venezuelan Oil Field Adventure

Coinbase co-founder Fred Ehrsam is venturing into the oil fields of Venezuela, a surprising shift for a prominent figure in the digital asset space. Through his company Primavera Infinita, he recently secured a production contract for the Budare-Elotes block with Venezuela's state oil firm PDVSA. This move into a politically volatile, sanction-scarred country highlights a bet on high returns from its reopening under new leadership and shifting U.S. foreign policy. Ehrsam’s investment reflects a venture capital-style appetite for risk, targeting assets deeply discounted by political uncertainty. He is not alone; smaller, politically connected U.S. firms like Aspect Holdings and Hunt Oil are also entering, while established giants like ExxonMobil remain cautious due to past expropriations. To manage the complex, capital-intensive nature of oil, Ehrsam is assembling a professional team. This trend extends beyond Ehrsam. Other crypto wealth, like BitMEX's Arthur Hayes and Tether, is diversifying into traditional hard assets—energy, metals, and agriculture—seeking physical scarcity as a long-term anchor. Tether, for instance, took a controlling stake in agricultural giant Adecoagro. These moves signify crypto capital's evolving interest: not just tokenizing real-world assets (RWA), but directly acquiring and operating them. Ultimately, Ehrsam's gamble is less on oil geology and more on the duration of Venezuela's current political window. It underscores a broader narrative where digital-era wealth seeks stability and scale in the physical world's most traditional, immovable resources.

marsbit3h ago

Crypto's Nouveau Riche Strikes Gold in the Real World: Coinbase Co-founder's Venezuelan Oil Field Adventure

marsbit3h ago

Refuting the Ethereum 'Abandoning' ETH Narrative: What Does It Really Mean to Pay Gas Without ETH?

Title: Refuting the "Ethereum Abandoning ETH" Argument: What Does Paying Gas Without ETH Really Mean? The debate sparked by Vitalik Buterin's discussion of EIP-8141 (Frame Transactions), which suggests users could pay transaction fees without holding ETH, has led to extreme claims that ETH will lose its value. However, this perspective misunderstands the proposal. Currently, an Ethereum user initiating a transaction must also pay the network's Gas fee in ETH. EIP-8141 aims to decouple these actions. It allows a transaction to be split into separate "frames." A user could sign a transaction to, for example, send USDC, while a separate Paymaster account pays the required ETH Gas fee on their behalf. The user would then settle the cost with the Paymaster using USDC or another token. From the user's perspective, they pay in a stablecoin without interacting with ETH. Crucially, from the Ethereum protocol's perspective, the Gas is still paid in ETH; only the settlement layer between the user and Paymaster changes. This concept isn't entirely new; ERC-4337's Account Abstraction already allows similar Gas sponsorship. EIP-8141 seeks to integrate this capability more natively. The core goal is to drastically improve user experience by abstracting away the complexity of Gas, similar to how one pays with a credit card abroad without handling the local currency. It also enables atomic operations, like bundling token approval with a swap, which would revert together if the swap fails. Regarding ETH's value, the argument that "no ETH is needed" is incorrect. While users may not hold ETH, Paymasters and services must still acquire and spend ETH to pay network fees on the backend. The demand for ETH shifts from being distributed across millions of user wallets to being concentrated in the balances of these service providers. The key variable is whether this improved usability attracts significant new users and increases overall network activity. If it does, total ETH burned in fees could rise substantially. If it doesn't, the change merely reshuffles who holds the ETH needed for Gas. In summary, EIP-8141 aims to lower the entry barrier by hiding Gas complexity, betting that this will drive broader adoption and increase the fundamental utility—and thus demand—for the Ethereum network and ETH itself.

marsbit3h ago

Refuting the Ethereum 'Abandoning' ETH Narrative: What Does It Really Mean to Pay Gas Without ETH?

marsbit3h ago

Trading

Spot
活动图片