Eric Lindauer

𝗔 𝘀𝗺𝗮𝗹𝗹 𝗯𝘂𝘁 𝘀𝗶𝗴𝗻𝗶𝗳𝗶𝗰𝗮𝗻𝘁 𝘀𝗵𝗶𝗳𝘁. Kraken has quietly launched FX perpetual futures—24/7 trading on pairs like EUR/USD and GBP/USD, built with the same infrastructure as crypto derivatives. No expiry. No market close. Just continuous access, the way crypto traders are used to. It’s a modest announcement on the surface, but it hints at something larger. Over the next decade, if stablecoins continue to gain traction and more trading flows through crypto-native rails, the architecture of markets could change. Not overnight, and not loudly—but gradually, and maybe irreversibly. What we consider “traditional” market structure—fixed hours, multi-day settlement, siloed asset classes—might give way to something more fluid, faster, and more unified. We don’t know exactly where this leads. But moves like this suggest the future of trading might not come from the places we expect. #Crypto #TradFi #MarketStructure #Stablecoins #Exchanges #Innovation #Kraken #FX https://bit.ly/4iubK9g

22

Some observations: we had an HFT prop firm reach out to us to get help hiring quant traders. They supposedly are a top 10 crypto market maker - they run quite stealth... but before they were willing to have a call with us, they wanted us to sign an NDA which included a 2 year non-solicit of their employees... Now I don't have a problem signing an NDA, but signing a non-solicit for 2 years with zero context and no guarantee of doing business (let alone us wanting to do business with them) is a non-starter. We asked them if they'd be willing to share more context on how many other recruiting partners they are using or plan to use, what kind of volume of hiring they anticipated, and if they would pay our fees. (We told them we were expensive). Instead of having a conversation, they told us that we were too expensive and that they were insulted that we would even request such a number, and based on that they would not be interested in doing business with us. Hmm. What's the point of this story? 1. We work for money just like everyone else. Companies that pay us well get our (and our team's) attention. 2. There are only so many firms we can work with in trading. We have to decide as a company who we hire for and who we hire from. We can't work with everyone. 3. If a company thinks of us as just a commodity, not a partner; or if a company can't understand or relate to our incentives to work with them - then trouble ahead. We're on a consolidation kick at Fionics and are really leaning into the firms we work with that see the bigger picture with us (and that means some companies might not make the cut), and that's just the nature of this business. If we're going to take on someone new to hire for, they need to meet our bar, just as much as we need to meet theirs.

60

7 commentaren

I spoke with a Head of Power Trading yesterday at a Prop Firm and they said they’ve not had much success hiring Senior Traders in the past. “They never seem to outperform the Traders we have trained ourselves” This isn’t the first time I’ve heard this either. When hiring Senior Traders it’s really important to not just focus on their PnL numbers and VAR allocation. You need to ensure you drill down on what contributes to their performance, how they trade and why they’re wanting to leave their current seat. It needs to be compatible with your environment. What tools do they have available? Do they have support from an analyst who builds the stack models or can they do this themselves? If you don’t have the models they need do you have the resources to build them and do you know how this will take? What market access do they need? What data do they have available? Do they work with a Meteorologist? There is never an exact science to hiring Senior Traders but what gives you a higher chance of success is hiring Traders from Firms that have a similar or worse set up to yours but still delivered good results especially in these market conditions. If you don’t have all the resources and set up of a Hedge Fund then don’t try and hire from that talent pool. If you’re a small to mid sized Prop Firm that expect Seniors to be able to build their own models then you need to hire Traders that can do this. In F1 terms if you hire a prime Lewis Hamilton but don’t have the car or team he had at Mercedes then don’t expect the same results. #HiringSeniorTraders #PowerTrading #NordSearch

63

4 commentaren

There are lots of ‘AI Engineer’ roles emerging now, and lots a grey in terms of what it actually means/what they actually will do Within the hedge fund world, this is broadly what they’re looking for Caveat - funds tend to vary in culture/role specifics so there is no true uniform answer. But broadly speaking, what does AI Engineer mean in a Fund? Multifaceted role comprising software engineering, data science and research. Generally work is split across creating AI models, building AI infra and developing AI tools. Some roles will naturally skew more in one direction than another, but most are a good mix of all three Generally there is a shared responsibility between engineers responsible for technical integrity of work/models and thebn PMs/Risk to use those models to generate PnL. They don't see ML as a guarantee of investment success (yet) Typically these roles sit in central platform teams and build core infra/data platforms, compute and shared ML tooling. The pods all build from these foundations. Some larger pods will want their own AI specialists but this would typically require specific asset class/trading knowledge also. Typically the weighting is on central Broadly speaking, AI Engineers are building platforms, models and tools to support Alpha-gen, data processing at scale, especially in Alt/Unstructured Data and managing risk / stress testing portfolios. But the scope is massive and we are at the start from a use case perspective That being said, there is definitely an element of FOMO and hype. Gains haven't met investment level yet but most have committed to an AI-first approach and will continue to invest heavily in the right people and technology. The shift is likely that hiring in AI will become more intentional and less kneejerk, and funds will push back to hiring monster engineers / data scientists / researchers to pivot heavily toward AI/ML, rather than your ‘AI specialist’ Finally - these are projects are launching / roles being hired with a multi year outlook in mind. So there is a big reluctance against anyone that is overly jumpy Super cool time to be an engineer in this space, and by default a recruiter :) Brave new world

29

6 commentaren

🔄 Tick-to-Trade in <100ns? Only with FPGAs. 🧠💡 In High-Frequency Trading (HFT), the holy grail is building an end-to-end tick-to-trade pipeline that can go from market data arrival to order generation in sub-microsecond latency. That’s only possible when you eliminate software bottlenecks — and FPGAs are the answer. 🎯 Here’s a simplified pipeline: [ Ethernet MAC ] ↓ [ Market Data Decoder ] → [ Order Book Logic ] ↓ ↓ [ Risk Checks ] [ Strategy Engine ] ↓ ↓ └─────▶ [ Order Builder + Network Transmit ] ⚙️ A Verilog snippet for order generation logic based on a bid price threshold: verilog module strategy_engine ( input wire [31:0] bid_price, input wire [31:0] target_price, output wire send_order ); assign send_order = (bid_price >= target_price); endmodule 📈 This logic can sit inline with a 10Gbps feed, and combined with real-time book updates and risk filters, can place orders before most firms even parse the packet. 💬 If you're building low-latency infrastructure or are curious about how firms push beyond software, let’s chat. #FPGA #HFT #TickToTrade #LowLatency #Quant #Verilog #FinanceEngineering #ElectronicTrading #NetworkDesign #MarketData

14

cheap dev now, valuation haircut later a £30k offshore build feels efficient until diligence opens the repo and the questions shift fast: provenance, licenses, reused modules, api exposure, real ownership. what looked like savings starts reading like risk. hidden technical debt compounds silently and erodes enterprise value. technology risk is a common trigger for price renegotiation. translation is simple: buyers rarely debate. they haircut the valuation. code is the asset. treat it like one before someone prices the doubt in. 💂 mechanics of value creation.

8

4 commentaren

Scale-ups, Big-tech and Hedge Funds are shunning junior engineers. With the total number of software engineer job openings way down on February 2020, this data shows that they're choosing one distinguished engineer over seven trainees. This chimes with our experience of the market. Over the last 3 years, our clients have been indexing heavily towards hiring experienced Talent Acquisition professionals into IC delivery roles, as opposed to trainees. Presumably because they believe it is them who can make the difference in securing these top-of-market engineers. Leaning on their superior market knowledge and potentially existing candidate relationships. It's a tough time to be starting a professional career. With AI showing so much potential, businesses are hesitant to hire trainees whose roles could be more rapidly usurped. Wage budgets are flowing to highly experienced individuals who can operate within the gaps and nuance where AI struggles to play. Largely judgment and taste. 

24

3 commentaren

𝗛𝗲𝗱𝗴𝗶𝗻𝗴 𝗹𝗼𝗼𝗸𝘀 𝗲𝗮𝘀𝘆 – 𝘂𝗻𝘁𝗶𝗹 𝗶𝘁 𝗶𝘀𝗻’𝘁. On paper, it’s simple: You’ve got an asset. There’s a risk you want to neutralise (or are required to because of regulation). So you slap on a hedge – probably a derivative – and call it a day. Sorted, right? Let’s take a common one: You’re managing a USD equity portfolio worth $100 million. Your base currency is SGD (becasue you are a Singapore based Instiution). So now you're long USD equities 𝘢𝘯𝘥 long USD vs SGD. To hedge the FX, you sell USDSGD forward – usually on a 3-month roll via FX swaps. In theory, your hedge notional tracks the portfolio's USD market value. Start at $100m. Three months later it’s $103m? You update the hedge. But theory starts to wobble when: • The hedge isn’t marked properly • MTM swings are ignored • Entry cost gets confused with full-cycle P&L • The hedge gets treated like a standalone trade instead of protection I’ve seen PMs brag they “saved 8bps” by waiting for a tighter spread – unhedged for days in the process. Legged, exposed, and completely missing the point. They celebrated a cheaper entry... while the hedge P&L was quietly flapping in the breeze. Then there's the  confusion betweern 𝘵𝘳𝘢𝘥𝘪𝘯𝘨 and 𝘩𝘦𝘥𝘨𝘪𝘯𝘨. The moment your hedge isn't in place – your're trading , not hedging. Fine if you are a prop trader. Bit of a problem if that is not your mandate. Now that’s before we even talk about 𝘮𝘢𝘯𝘢𝘨𝘪𝘯𝘨 the hedge mid-cycle: Taking profits. Restructuring. Exploring options or hybrids. You know… optimising. Outside a bank market makers trading book, most hedges are treated like fire extinguishers – set and forgotten until something explodes. That’s risky enough in FX. But with long-dated interest rate swaps? It can be a slow-motion disaster – especially for insurers who think “set and forget” is a strategy. But a good hedge doesn’t just sit there. It earns its keep. 💬 What’s the worst hedge mistake you’ve seen (or made)? No judgment – only stories. #mikeduncan #ParaBellumAdvisors #ᴍᴀᴋᴇitʜᴀᴩᴩᴇɴ #FXHedging #Derivatives #Hedging #RiskManagement #CapitalMarkets #Structuring   #LBFalumni #SkyHighTower🚀

44

7 commentaren

He stumbled on the logic question... Messed up the data structure. Admitted he’d never coded in Rust. I thought it was over. But here’s what he did differently: → Asked sharp questions about the trading model → Suggested how to tighten their signals → Built instant rapport with the PM The fund didn’t care that he bombed one round. They saw potential. This industry doesn’t hire perfect. It hires sharp. If you can think like a trader, add alpha, and hold your nerve in the room… You’ve got more edge than any polished CV can show.

17

Mosaic vs. Eurex case isn't just a technical spat. It's a warning shot for every trading venue, broker, and liquidity provider still pretending that nanoseconds don’t matter. Mosaic claims they lost €75 million because competitors gamed Eurex’s frame-per-second limit using malformed ethernet headers. We’re talking 3.2 nanoseconds of edge. That’s how far this game has gone. They even gave it a name: “Corrupted Speculative Triggering.” It’s not just clever. It’s corrosive. Eurex denies wrongdoing, calling their version “Advanced Speculative Triggering” and saying everything stayed within the rulebook. But here’s the kicker: their rules don’t even count preambles and SFDs toward frame limits. So technically, it’s legal. Ethically? That’s another story. What this reveals is a new layer of market asymmetry, not just between HFTs and retail traders, but between those with legal engineers and those without. And regulators? Mosaic’s first attempt to intervene got blocked by a court that said “this isn’t about you, it’s about the public interest.” They’re now knocking on ESMA’s door. But the damage to confidence is already done. If you're a forex broker or liquidity provider, don’t dismiss this as a derivatives issue. The exploit is technical, not asset-specific. It’s a blueprint for asymmetric warfare in any electronic market. Here’s what’s at stake: - Nanosecond advantages are no longer theoretical, they are monetizable. - Surveillance systems might be technically robust yet conceptually blind. - Exchanges are setting latency caps that can be circumvented with legal packet tricks. - Regulators may be under-resourced or procedurally constrained. - Fair access is becoming an illusion unless protections like PLP (Passive Liquidity Protection) become standard. For AI builders and traders, this opens both a threat and an opportunity. AI can detect latency arbitrage, optimize routing, and even defend against toxic flow. But it can also be weaponized to exploit every technical edge unless guardrails are in place. So the bigger question isn't whether Mosaic is right or Eurex is wrong. It's whether we’re ready for a market where milliseconds are slow, and compliance is 5 years behind the tech curve. - Invest in AI-driven toxic flow detection and nanosecond timestamping, or partner with Liquidity providers that are knowledgeable in this. - Push for latency-neutral venues, not just faster feeds. - Don’t rely on public regulators to guard your business edge. - Stay ahead by designing for fairness, not just speed. What’s your defence plan when someone beats you not with better trading, but with a better packet? #HFT #Forex #AI #LatencyArbitrage #AIFinance #LiquidityRisk

10

2 commentaren

There’s a quiet hiring wave happening across crypto funds — but it’s not the “growth at all costs” type we saw in 2021. Today’s mandates are precise. Funds want PMs who can show live, audited performance and developers who can build real trading infrastructure. The days of “proof of concept” hiring are over. We’re seeing the most traction for: Systematic PMs with strong Sharpe and scalable capacity. Developers with low-latency and exchange integration experience. Teams that treat operations and risk controls as alpha drivers. The bar has gone up — but so has opportunity for those who can meet it. #CryptoRecruiting #FundManagement #DigitalAssets #SkillTree

21

1 commentaar

"Canva shares trade on US platforms despite explicit company restrictions" This article, which appeared on Capital Brief last week (article link in Comments), details how shares of Canva are appearing on US trading platforms, despite the company not allowing direct secondary transactions. It is generally easy to figure out whether companies allow for direct transactions; so why do trading platforms allow these listings if they know they won't be approved? Not commenting on this particular deal, but taking a more general view, if the intention is to force the hand of the companies, that are reluctant to allow these transactions, to loosen their policy, for the benefit of their employees, then this effort is near-sighted and problematic. When unapproved bids, or exaggerated terms, are listed, only to be found out through due diligence that these deals are not as advertised, major #Retail and #Institutional players may be reluctant to participate- we hear this complaint time and time again in our conversations. For #PrivateMarkets trading activity to truly take off, major buyers need to step in, similar to Nancy Pelosi's $5M secondary transaction of Databricks on the Forge platform in March, 2024. A platform with real-time data, transparent bids, with an intuitive UX, which is what Forge's Next Generation Marketplace is aiming to accomplish, will hopefully bring more of a robust infrastructure to these types of deals. Only then will serious players step into the arena and engage in large-block trades with confidence. Equity Shift, Addepar, Augment, Hiive, EquityZen, Nasdaq Private Market, Carta, Christine Healey, Stableton, Manhattan Venture Partners, Caplight, Pulley, Monark Markets, AngelList, Equitybee, Linqto, Nasdaq, Destiny (D/XYZ), Collective Liquidity, CAIS, iCapital, Zanbato, Sydecar, Percent, EQUIAM, Delio, Qapita, Eqvista, InvestX, StartEngine, Long-Term Stock Exchange , Ledgy, Moonfare

12

2 commentaren

Asymmetric idea (2) - MSTR MSTR is a very complex beast. Weirdly, it is one of the most loved and most hated tickers. Most people write off MSTR in bear markets and this time the noise is stronger. Intuitively, I don't like MSTR. I don't like that a free, beautiful asset like bitcoin is cornered and accumulated by a single corporate entity. Historically, I have studied the fate of amalgamated copper and hunt brothers, and it should not be any different for MSTR. But I can't ignore that it could also present an asymmetric financial opportunity in the near future. My thesis is simple - MSTR stock will fall to USD 20-40 (down 95% from its peak of 543 and it is already down to 120s) at the bottom of this bear market (sometime in Q4). At 30-40K USD btc price - market cap would be roughly equivalent to btc value - debt (including preferred) - some discount (the spring will compress due to sentiment) MSTR won't get liquidated as the debt burden is still manageable although the sentiment would be that of "insolvency" And at the peak of the next cycle, MSTR will have more than 1 million btc, and assuming bitcoin price of USD 250,000 and NAV multiple of 2 (this cycle and previous it went as high as 4+) it should be a 500 billion equity valuation so a 50X from bottom this year. Buy end of 2026, sell end of 2028. 2 years, a potential 50x. Stay patient this year. Things will be available cheap. Generational wealth creation opportunity may be on the horizon. Again, do your own research. These are just general ideas. Not financial advice. SatoshiWealth

62

7 commentaren

The US HFT duopoly just became a three-way arms race. Citadel Securities and Jane Street have dominated US algorithmic trading for years. Predictable competition. Known quantities. Now? Optiver is coming to Manhattan. Three distinct trading philosophies competing for the same microseconds. The talent wars just intensified. Compensation packages are climbing. Technology development accelerated. But there's something deeper happening here. → European risk models vs American execution → Different cultural approaches to algo trading → Fresh strategies disrupting established patterns Jane Street's mathematical approach. Citadel's scale advantage. Optiver's market-making expertise. Each brings something different to the table. For trading professionals, this creates career opportunities that didn't exist six months ago. New paths. Better compensation benchmarks. Alternative cultural fits. The comfortable two-horse race? Gone. Who do you think wins in this three-way battle?

20