The Proteus test: Q126 letter to partners
Dear partners,
Year-to-date as of March 31 2026, Stone Sentinel Capital (“SSC”) lost 7.8% while the S&P 500 Index (“S&P”) declined 4.3%. Figures are gross returns in USD. Year-to-date returns are not annualized.
While YTD returns have been unsatisfactory, your manager has little control over the market prices of securities.
Your manager only has complete control over what securities to own.
The future is uncertain. A security can trade at any price.
But SSC is positioned to benefit from and guard against the uncertain future.
Risk is not uncertainty
Not all unknowns of the future are equal.
The economist Frank Knight differentiated between risk and uncertainty.
Risk relates to unknowns in which outcomes can be assigned probabilities.
Within risky unknowns, the nature of the future lies within expectations.
Uncertainty describes unknowns in which outcomes cannot be imagined and probabilities cannot be applied.
Within uncertain unknowns, the nature of the future is unexpected.
To illustrate, the outcome of a coin toss is risky, but that of a world war is uncertain.
Risky and uncertain problems also have different solutions.
A risky problem is solved by understanding bounded outcomes based on probabilities and history.
But an uncertain problem does not have history.
Without history as a guide, uncertainty presents unbounded outcomes.
Proper positioning is the solution to unbounded outcomes.
If you go through a war and don’t know when it will end, you should conserve food, water, and other basic necessities.
Saving as much resources as possible to weather the war is wise positioning. It is unwise to guess when the war would end and optimize saving-consumption mix until then.
The example illustrates that uncertain problems cannot be solved with risk-solutions.
The solution must match the nature of the problem.
Investing may map neatly onto Knight’s framework.
Investors may play the fundamental game of risk. Grounded by the gravity of fundamentals, they emphasize history and concrete evidence that point to an expected range of outcomes.
They may also discard the fundamental game of risk and be engaged in the sentimental game of uncertainty instead. These investors rely on intelligent forecasts of the future that have few parallels with history. Without history as a concrete guide, outcomes vary widely and change quickly based on prevailing sentiment.
To illustrate, investing in artificial intelligence (AI), which has little to no parallels in history, appeals more to investors of the second than the first nature.
As discussed, an uncertain problem cannot be solved with a risk-solution.
Similarly a game of sentimental uncertainty cannot be won with a fundamental-risk solution.
Investors in AI rely much more on imagination and sentiment than on history to assess outcomes.
Drawing parallels of AI to prior technological advances or other history is foolhardy from the start. One should not rely on history when there isn’t any.
Each game demands unique strategies that are inapplicable to the other.
To win is to pick the right game first, then devise the appropriate strategy.
Your manager spends as much effort picking the right securities to analyze as he does in analyzing their underlying businesses.
Because the future of most things is uncertain, your manager firmly elects to play on the fundamental court of risk, and engages only in businesses whose futures are risky rather than uncertain — businesses where outcomes can be bounded and assessed.
Your manager strives to avoid mistaking uncertainty for risk.
The Proteus Test
Your manager has no insight as to when and how two wars, disruptions in energy supplies, and the unknown impacts of AI will resolve, but is reminded of a story that may shed a little light on what to do.
In The Odyssey, Menelaus (king of Sparta) recounted being stranded on the island of Pharos on his way home from Troy.
He needs to know how to get home but only Proteus, the old sea god who herds seals on the beach, has the answer.
The problem is that Proteus will only speak truth if physically restrained. The moment anyone grabs him, he shape-shifts violently through every form to escape.
Proteus's daughter Eidothea takes pity on Menelaus and tells him the trick:
You have to hold on through every transformation, no matter how terrifying, and simply not let go.
When Menelaus and his men ambush Proteus, the sea god becomes a lion, then a serpent, then a leopard, then a boar, then running water, then a tree.
Menelaus holds on through all of it. Eventually Proteus exhausts his forms, returns to himself, and answers Menelaus's questions truthfully.
The wars, energy disruptions, and AI panic are akin to Proteus’s transformations if they have little bearing on the underlying businesses of stocks.
They are the price of reward.
Menelaus holds through terror for answers. Investors must endure volatility for rewards.
While Menelaus is certain about Proteus’s answers, investors ought to be sure that they are holding the right things.
Your manager confirms by underwriting positions again with a special emphasis on the macro events underlying volatility.
The point of picking securities with limited downside and significant upside is to ensure few ways to lose and many ways to win.
When done right, positions are well-defended against the vicissitudes of the market.
However, excellent defense loses all meaning if the owners abandon ship for the wrong reasons.
It’s one thing to jump ship because of a breach in defense.
It’s another to do so because of a loss of bearings from sea-sickness.
Investors must strive to practice the former and avoid the latter. It's rational to sell when the thesis is broken, but destructive to do so because the drawdown is uncomfortable.
Portfolio update
SSC is currently invested in 4 stocks that are listed in public markets in Hong Kong, Malaysia, United States, and Japan.
The two primary contributors to the drawdown are Ascentech and Protasco.
Ascentech is the Japanese operator of Citrix, one of two leading global VDI (virtual desktop infrastructure) providers. The stock declined about 30% from its 2026-high because of fears of AI disruption and weak guidance.
The “SaaS is dead” narrative concerns AI displacing software incumbents by favoring AI-native software or making software so cheap to build that customers build their own.
But the narrative is about application-layer disruption. Ascentech operates at the infrastructure layer — the plumbing that sits beneath all of these applications.
Whether a company uses Salesforce or an AI-native CRM alternative, they still need what Ascentech sells — secure remote access for their employees, zero trust security for their data, and VDI or thin client infrastructure for their endpoints.
If anything, the AI-disruption thesis increases demand for Ascentech products. As organizations adopt AI-native alternatives, they're adding complexity to their IT environments with more applications and cloud services. That complexity drives demand for Ascentech’s products from VDI to DLP (Data Loss Prevention for protecting data as it moves between more applications) and hardware necessary to run AI securely on-premise.
As for guidance, next-year guidance looks weak because the company guided for flat revenues and lower margins (~16% in 2026 to ~11% in 2027). These seem alarming on the surface but aren’t so because revenues and margins were inflated in 2026.
A surge of customers in 2026 from a convergence of industry drivers (Windows 10 EOL, VMware migration, financial institution hardening) led to large upfront payments for deployment that are typical for Citrix contracts. The upfront payments will not be repeated and customers are responsible only for license fees after deployment. The high-margin nature of the upfront payments inflated margins as well.
Ascentech earns 20-25% ROE, and trades at 12x 2027 earnings and 8.4x 2028 earnings from guidance, supported by 46% insider ownership including founders, the CEO, and a SoftBank-affiliated PE fund that bought its stake at a price roughly 40% higher. Dividend payout has been raised twice in the past year, and roughly 76% of its market cap sits in cash.
Management has never missed guidance and usually beat-and-raise. Guidance should hence be treated as the baseline.
Protasco is one of 3 companies that maintain roads under federal contracts in Peninsular Malaysia. The stock is down 21% from its 2026-high.
The decline is likely caused by the uncertainty of its most important “Roadcare” contract. Protasco owns 51% of the contract, which accounts for 60% of revenues. The contract is currently in a 1-year bridge agreement, so it is uncertain as to whether the contract will be extended past February 2027 (expiry of bridge agreement).
The odds appear overwhelmingly in favor of an extension. First, history is on the side of shareholders.
No federal road maintenance incumbent has ever been displaced through competitive tender in Peninsular Malaysia. Roadcare has also operated for 25 years with no performance-related termination or public criticism of service quality.
The second-largest owner of Roadcare after Protasco is KOP Mantap, wholly owned by the Royal Malaysia Police Cooperative. The contract benefits the police union directly and may face political friction if not extended.
The smallest owner is Muhibbah Engineering, which is a public company 3x the size of Protasco with a 50+ year track record and no history of government contract cancellation.
All three partners have been aligned for 25 years.
Second, switching costs are significant. A new entrant would need to replicate Roadcare's depot network, road inventory data, monitoring system integration, relationships with JKR district offices across four states, and the supervisory capacity of ~380 sub-contractors.
The only road maintenance incumbent to be nearly displaced is CMS. CMS was awarded a 15-year maintenance contract concession in 2003. But the renewal in 2019 came with reduced scope and ownership.
However, the outcome may be deserved because it belonged to the Taib dynasty, which was the subject of massive international corruption allegations.
If politically-toxic CMS that’s most deserving of displacement was only punished with reduced scope, it’s very unlikely that a functioning private-public consortium will be met with displacement.
But the market appears sure that Protasco will be displaced by punishing its stock to the point of 1.3x trailing earnings (not a typo) and 0.2x book value.
This is not all. Net cash is now roughly 2.5x market cap. And Protasco has land on the balance sheet that was last appraised in 2002 and likely worth substantially more at current market values.
The downside appears well-protected with potential for significant upside after the uncertainty of the concession is cleared.
Your manager welcomes questions and comments at mg@stonescap.com.
At your service,
Marcel Gozali
