Vinod Krishnan

FDE Partner @Ekimetrics Data & AI London | Value Investing meets AI meets Macroeconomics

[Vinod Krishnan @vknotkrishnan]

https://t.co/E5c7w5ZRBN

[Vinod Krishnan @vknotkrishnan]

Love this! Congratulations to team @PromptQL - resonate with a lot of these lessons!

[Tanmai Gopal @tanmaigo]

https://t.co/tYLNPuaVi4

[Vinod Krishnan @vknotkrishnan]

Phenomenal interview! Martin’s really one of the GOATs.

[TBPN @tbpn]

FULL INTERVIEW: @MartinShkreli breaks down the collapse of Leopold Aschenbrenner's Situational Awareness, and how Citadel capitalized on the liquidation.

01:00 Why Situational Awareness collapsed
03:00 SALP's 4x leverage, firms that were tapped to buy
08:00 Why Citadel did the deal
12:00 Where Leopold went wrong
14:30 Why AI funds could still be at risk
20:15 Why you can't just hit 'Sell' on a huge position as a hedge fund
26:15 Ken Griffin and Citadel
30:00 Can Leopold rebuild?

[Vinod Krishnan @vknotkrishnan]

For most people, selling labor is how wealth begins. As you climb the ranks, you accumulate capital - allocating it across cash, real estate, stocks, or perhaps even crypto. Index funds tracking the S&P 500 or Nasdaq have become the default choice for retail, leaving most pensions and IRAs heavily exposed to equity returns.

The problem? Many of these accounts are now systemically overexposed to high valuations driven by animal spirits from the AI super-cycle. When retirement portfolios hold trillion-dollar companies trading at 100x sales with negative earnings, alarm bells should be ringing. As @michaeljburry notes, 95% of people have no idea what they’re actually investing in! He's not wrong.

The case for being far more deliberate with your capital has never been stronger. Yet, most people lack the time or patience to do the heavy lifting.

If you are willing to put in the work - going against hot momentum trades, building competence in familiar sectors, and staying patient - you can compound wealth in vastly safer ways:
> Risk is not volatility; risk is the probability of permanent capital loss
> The higher the price you pay, the higher your risk of underperformance
> Buy quality businesses at deep discounts, and you set yourself up for outsized returns

This investing research tool - built on Claude Co-work with a surprisngly easy connection to Nemotron ($NVDA's open-source model) - serves one core purpose: helping you figure out where to look.

[Vinod Krishnan @vknotkrishnan]

https://t.co/E5c7w5ZRBN

[Vinod Krishnan @vknotkrishnan]

@rajoshighosh Thanks Jo! Much appreciated! Looking forward to building this out in much more depth, potentially on @PromptQL - to make life easier on the underlying models, especially if/as this gets adopted by investing teams that need shared context! What a time to be alive!! 🚀

[Vinod Krishnan @vknotkrishnan]

@Jason A Value investing research agent using Fable (for design, maintenance and escalations) and Nemotron (for scale) - built with Claude Co-work. More here: https://t.co/a4NTZlAHAe

[Vinod Krishnan @vknotkrishnan]

https://t.co/E5c7w5ZRBN

[Vinod Krishnan @vknotkrishnan]

“While the great innovation that comes from model providers having fair use rights to train models on public data is needed, I find it ironic that the status quo is to then turn around and impose restrictive terms on distillation, and to reserve the right to learn from customer usage and interaction data.”

Spot on!

[Satya Nadella @satyanadella]

https://t.co/xv6csf1SbV

[Vinod Krishnan @vknotkrishnan]

Great list. What I’d add to this list is that markets will continue to be almost always inefficient and mispriced. The patient, emotionally mature investor will see another day when 1$ of value trades at 50c, in a space she/he understands well. Generational returns aren’t the territory of only VCs/private investors :)

[Amanda Orson @amandaorson]

The faster technology moves, the more I think about Bezos' question

What won't change in the next 10 years?

Things I've been writing down over time:

- Humans will always need shelter, food, energy, and healthcare.
- The desire for ownership and the accumulation of wealth.
- The physical world will move more slowly than the digital one.
- Every increase in technological capability, especially AI, will require more energy.
- People and businesses will continue to need access to capital.
- Capital will continue to seek returns that exceed inflation.
- Underwriting methods evolve, but demand for credit (loans) is persistent.
- Trust remains scarce and becomes increasingly valuable as content, code, and fraud become cheaper.
- Verified identities and reputation becomes more important as information becomes abundant and synthetic.
- Long-term wealth creation and dynastic (multi-generational) thinking predate modern technology, and will persist.
- Coordination and transaction costs never fully disappear; market friction will continue to justify the existence of firms and intermediaries.
- People will continue to compete for status.
- Consumers will pay a premium for products and services that confer status.
- Time remains fixed at 24 hours per day.
- But attention is a finite resource and an enduring constraint.
- Products that credibly save time (or enable delegation) have a perpetual market.
- Inaccessible, proprietary data will be a persistent moat. The more inaccessible and difficult to aggregate, the deeper the moat.
- People want accountability, recourse, and clearly identifiable responsibility when things go wrong.
- Regulation consistently lags technological innovation.
- Compliance requirements, licensing, and regulatory moats persist even when machines can perform the underlying task.
- Local knowledge remains valuable and difficult to replicate.
- Heterogeneous markets (like real estate) continue to reward people with deep contextual understanding.
- Incumbent organizations tend to underinvest in disrupting their own businesses, which always creates opportunities for challengers.

Bezos' insight on what wouldn't change in 10 years was "Customers will always want lower prices and faster delivery."

It's boring/ true, but I think that's the point.

Everything we build today can and will be rebuilt more cheaply, faster by someone else.

Build on the invariants, not the trends.

What have I missed?

[Vinod Krishnan @vknotkrishnan]

I just read Cembalest's "Semiquincententacles" (JPM Eye on the Market) in full. The thesis: on America's 250th birthday (tomorrow!), the US grip on global markets is still iron-tight - but the actual risks lingering are getting less air time than perhaps they should!

*Reserve currency:* the dollar ignored its own obituary. Despite 125% debt/GDP and record sanctions, all six reserve-status metrics (FX reserves, invoicing, SWIFT, cross-border loans, etc.) are stable. Gold's "surge" to 29% of reserves is almost entirely a price effect, not central banks dumping dollars (though I'd question if he's accounting for undeclared gold purchases!!). And the CNY as a reserve currency - he says that is preposterous right now (and I agree) - $60T in banking assets (~50% of global GDP), $800B–$1T in 2025 capital outflows, money creation outside the central bank's control. One does not run a reserve currency behind a closed capital account.

"Sell America" fizzled. After the March tariff scare, the trade-weighted dollar is down just 1%, US yields rose less than the rest of the DM world, and US equities outperformed. Long term, US stocks have crushed RoW even after the non-US recovery - because in nearly every sector US firms earn higher ROE/ROA. The P/E "discount" abroad has been a bug zapper for allocators.

AI is the whole ballgame - and it is also the risk. The AI basket has driven ~75% of S&P returns/profits/capex since Q4 '22. Warning lights flashing: semi technicals at dot-com extremes, leveraged-ETF rebalancing impact up 5x, thin breadth, margin debt climbing. But memory might be turning from cyclical to structural - with possible rerating on the cards for the MU's of the world.

For AI builders, the most interesting signal: the moat is leaking. NVIDIA's share of accelerator revenue is declining as ASICs (Trainium, TPUs) improve and inference overtakes training. Frontier labs are raising token prices while open + Chinese models (DeepSeek, Qwen, Kimi, GLM) land within a few Elo points at a fraction of the cost. And as has been the talk of the town lately, agent harnesses (memory, tools, safety boundaries and orchestration layers in Claude Code) that can run open models often improves their output quality and reduces the need to rely on the most expensive frontier models. Exaclty what I've been tinkering with on my mac's local-hardware and getting my clients to pivot to! A one-time set up fee for higher privacy, lower cost and a mechanism that really forces you to think hard and deep about what you're trying to solve - and ZERO dependency on the "Fable-kill switch"!

The criticality and hence the sensitivity that is Taiwan. Semiconductor trade now exceeds crude oil trade. 8 of the 10 biggest companies on earth depend on TSMC (I'd argue even more) - and Taiwan is the most blockade-sensitive advanced economy alive (imports ~97% of energy, 60% of food, days of gas storage). China is closing the gap fast (Huawei's vertical chip design, rising GPU self-sufficiency) despite not being able to access NVDA/ASML's EUV lithography.

And the part most market commentary underweights - Cembalest's two biggest medium-term worries are the rising unpredictability of the rule of law and the defunding of US science - not debt or inflation!

Bottom line: betting against America has been a losing trade for 40 years and probably still is. But the eagle-octopus's grip depends on two things money can't easily buy back - credible institutions and a scientific edge. Those might trump the dollar bears. 🦅🐙

All of this has masive consequences for where to find value - ie where can one still buy a dollar for 50c. And I'd argue there's plenty of those lurking, depsite what might seem like ATHs everywhere!

Happy 4th of July to those celebrating!

[Joe Weisenthal @TheStalwart]

New Michael Cembalest report about the ongoing dominance of US financial assets https://t.co/Pu0IriyS2E

[Vinod Krishnan @vknotkrishnan]

https://t.co/qjQLSPSS6f

[Vinod Krishnan @vknotkrishnan]

Agree. Simplest version of the thesis:

- 90-95% of enterprise workflows don't need frontier intelligence - extraction, routing, summarization, boilerplate code. Open models nail it. Paying frontier token rates to run that at volume is madness.
- I'll go one step futher: Frontier is worth every penny on the hard 5%. Everything else is a barbell: cheap/open inference for scale, increasingly self-hosted - Qwen-class on high-mem GPU/CPU, inside the firewall.
- Same move firms made in the 80s/90s: commoditized capability gets pulled in-house. Labs keep the premium, lose the volume. There's your cap on pricing power.

[Tommy @Shaughnessy119]

The most basic way AI could blow up imo. I'm not saying it does but this is the most obvious way I can see it happening

- Per seat subscriptions are massively subsidized. The flat fee was priced way below what heavy usage actually costs

- For real business use you have to move to the API anyway. Data protections, work integrations and compliance officer approval

- On the API you pay metered rates, and businesses are burning credits way faster than the per seat pricing ever led them to expect

- This is everywhere right now. Internally for us, Codex users, Uber torching its entire 2026 AI budget in 4 months, the Microsoft comments. Just go try an API

I shared more on this here: https://t.co/iZrqrCAIRW

- And I don't think most businesses have the money to keep paying increasing API rates without a real change to how they operate (caps needed)

- Because they have a cheap alternative. They can reach open source models through any aggregator (OpenRouter, Venice, Baseten, Together) and still get strong privacy. Venice private data centers, or E2EE/TEE serving GLM 5.1.

More on open source inference provider raises here: https://t.co/7kf56P44yQ

- And the discount is enormous. DeepSeek V4 codes within a hair of Opus on SWE bench at roughly 1/30th the price, and the cheapest open models run closer to 1/100th

- Chinese labs open source frontier grade models. The model is the single biggest cost an inference provider has, and they get it for free

- This idea dies if China goes closed source. That is actually bullish web2 AI labs, because if everyone is closed you pay up for the best intelligence. China goes closed source if they are tired of giving away an asset and they want the revenue and data flow to train new models

- Is this showing up in web2 AI lab revenue yet? No. Revenue is off the charts. Anthropic went from 9B to 47B run rate in five months

- So go forward, what happens?

- I think revenue slowly starts leaking to the open source inference providers (see Venice usage, OpenRouter's $113M raise, Baseten is raising at $11B or triple its valuation in three months, on revenue that went from $200M to $600M annualized in a single quarter)

- It doesnt move overnight, but it caps the labs ability to raise prices, and margins are already deeply negative. OpenAI is reportedly running near negative 122%

- With margins that bad there is no cash flow, so the labs are fully dependent on outside capital to buy GPUs, train models, and keep subsidizing usage (I.e. see Google tapping $80b equity sale, granted 30b for employee RSU taxes. Clearly they think Equity is overvalued or you wouldn't sell it)

- The break comes when that capital stops. Pricing is capped so margins cant improve, and the moment investors lose conviction on payback, the whole flow reverses

- Why would they lose conviction on payback? Back to the start - the inability to improve margins or get businesses to pay more

- This is also limiting, if we start making new drugs with AI or create entirely new businesses, you better believe people will pay up to the max for AI usage

[Vinod Krishnan @vknotkrishnan]

https://t.co/1JFAXmNyvH

[Vinod Krishnan @vknotkrishnan]

second time now Jaya’s written something I wish I’d written!

[Jaya Gupta @JayaGup10]

https://t.co/5vDrndcmFn

[Jaya Gupta @JayaGup10]

https://t.co/5vDrndcmFn

[Vinod Krishnan @vknotkrishnan]

@ramahluwalia - what you were hinting at yesterday! :)

[Vinod Krishnan @vknotkrishnan]

Time for a stock pitch! $BR Broadridge Financial Solutions is the most boring stock you've never owned. And it is quietly the most compelling buy in the market right now. 🧵1/

[Vinod Krishnan @vknotkrishnan]

15/ Summary: $BR is a regulatory-moated, 40% ROE, near-monopoly compounder with record FCF yield, 55% upside to consensus, a 19yr dividend growth streak & a business that processes positions tied to $100T in global AUM. In a market where nothing feels cheap - $BR does. 🧵 END

[Vinod Krishnan @vknotkrishnan]

Time for a stock pitch! $BR Broadridge Financial Solutions is the most boring stock you've never owned. And it is quietly the most compelling buy in the market right now. 🧵1/

[Vinod Krishnan @vknotkrishnan]

14/ 3 risks but all manageable: 1. regulators shift to "notice & access only" for proxies - slow burn, not a cliff. 2. Big banks build in-house alternatives - economically irrational, never happened at scale. 3. Rates stay high, multiples compressed. None truly break the thesis.