I'm meeting new teams every week right now, and I want to start with the good news, because there isn't a lot of them lately.

A real wave of new quants is entering this space - and they're good. Genuinely institutional backgrounds, professional from the first call, PhD candidates and TradFi guys seeing opportunity.

They show up asking the right beginner questions: how long does my track record need to be, what AUM can I realistically get, what infrastructure do I actually need, which exchange should I be on, etcetc.

These are the right questions, and the fact that serious people keep asking them is one of the healthiest signals in the industry. Fresh blood, high quality, constant inflow. The universe 1 year ago was completely different than it is now. Good luck keeping up with it.

And at the very same time… some of the older teams are packing their bags and leaving.

The promised land took too long. The "American dream" version of the SMA business - the one where you build a clean book, raise steady AUM, and compound your way to a real firm - is STILL real, but the timeline to get there is brutal, and not everyone can outlast it. This is fine. The Law of Nature.

The reasons are the usual suspects: underperformance, the grind of raising capital in a dry market, and a four-year cycle that - surprise surprise - is happening yet again, unfolding in a way that's almost impossible to time and predict (just like the good old markets).

So new teams in, tired teams out. That's the door this industry spins on.

Let me talk about THAT ONE thing that decides which direction you're walking through it: underperformance - what it actually costs you, and when you should start to worry.

Here's the mechanic most teams don't understand until it's too late.

Allocators carry a buffer. When you sign a test account - say 100K SMA, the standard size before any scaling conversation - you get roughly three months of patience. If you don't perform in that window, it's fine. Genuinely fine. It's understood as noise. Nobody panics over a soft first quarter on a test ticket.

But push past that. Four months. Five months of flat-to-negative…

That's where the eyebrows go up. That's the zone where the allocator quietly starts deciding to pull the trigger aka withdrawing the capital, reallocate the exact same money to someone else. And then the cycle just repeats with the next team in line.

I've said it before and I'll keep saying it because it's the single most important sentence in this business:

There is no relationship. Performance is the relationship.

When the performance stops telling a good story, the relationship dies on the same schedule. The allocator has their own pressures, their own scorecard, their own people to answer to. Your soft five months becomes their problem upstream, and they solve it by moving on.

Now the part that really stings, and that teams consistently refuse to believe until they live it:

Once you lose an allocator to underperformance, getting them back is brutally hard.

They'll be polite about it. "Okay, let's stay in touch - let's sync again in three months and see where you're at." That sounds like a door left open. It is almost never a door left open. It's a courteous exit. What it usually means in practice is: we're done for now, and probably for good. Maybe a miracle will turn this around. Miracles exist, sometimes you hear someone winning a lottery, but it rarely is you.

And think about what that actually costs you. You don't just lose a ticket. You lose a relationship you already paid for in time and trust - and now you have to go find a brand-new allocator, in a market where allocators are scarce, start the trust-building from zero, run a fresh test amount, and perform exceptionally well from day one… all while carrying the quiet stink of "the last guy stopped working with them."

You're rebuilding the whole staircase from the ground floor.

Which is exactly why the move is to not let it get there.

For quant trading teams - the moment you sense an allocator starting to drift - act. Don't wait for the withdrawal email. Everything you do from your side in that window is cheaper than replacing them.

Over-communicate. This is the time to increase your monthly cadence, not go quiet and hope the chart fixes itself. Allocators remember the team that talked through the pain, not the one that reappeared when the equity curve looked pretty again.

Offer the one-on-one call. Get on the line, walk them through what's happening, what you're seeing, what you're changing and what you're deliberately not changing. Show them there's a live brain attached to their capital.

And consider a small gesture of intent — a modest perk, a fee accommodation, something that costs you little but signals clearly: I am invested in keeping this relationship. The intent has to come from the team side, visibly. A small concession at the right moment can buy you another month of patience, and another month is sometimes all you need for the strategy to turn.

Because the alternative - new lead, new relationship, new testing amount, and the obligation to outperform from scratch - is a far more expensive road than a few proactive calls and a small show of good faith.

One way to keep the performance steady is to constantly upgrade the tools you are using in-house.

Please welcome a new section of the newsletter where we showcase institutional products that will help quant teams to overdeliver.

The Aperiodic team spent years figuring out how to effectively ingest, store and query hundreds of terabytes of raw data, just to get the reliable, point-in-time aggregates for their MFT use-cases. It ended up providing them with the edge they were looking for.

They now decided to open it up for quant teams. 

100+ market microstructure, liquidity and flow metrics (slippage, book imbalance, flow toxicity, Kyle lambda, run imbalance, funding, basis), precomputed and served as immediately usable parquet files over an API. Point-in-time with no rate limits.

This helps quant teams to get from an idea to a clean feature matrix in a couple of commands, without standing up any of the infrastructure yourself.

Start with the notebooks: you can use the preview data without a subscription or contact Mark for a trial.

Now let me flip the table, because the allocators reading this make a mistake that's just as costly, and almost nobody talks about it.

Most allocators badly underestimate the variance baked into manager selection.

Here's the scenario I watch play out constantly. Take a category - directional, stat-arb, long-short, pick one. I introduce, say, seven credible teams in that bucket. The allocator studies the past performance, picks two or three for the portfolio, and feels good about a data-driven decision.

Six months later, the picks are flat. And here's the cruel twist: the teams they passed on are the ones printing. Quietly, asymmetrically outperforming the exact names that made the cut. (Seriously seen this countless times)

So the allocator comes to me frustrated - "why aren't my teams performing like they should?" - completely blind to the fact that the other teams from that same shortlist, the ones they evaluated and declined, are having a great run. They're scoring their decision against an empty bench instead of against the field they actually chose from.

There is a real luck element in this game - yes. Past performance over a short window is a weak predictor of the next six months, and pretending otherwise is how good allocators make avoidable errors. You're not as much in control of the outcome as the selection ritual makes you feel.

So how do you fight variance instead of getting fooled by it? Two things:

One - stay close to the whole field, not just your picks. Check in monthly with every team in the category, not only the ones you funded. Ask how the month went across the board. The information you need to correct a bad pick is sitting in the teams you didn't choose.

Two - and this is what we actually prefer - spread small. Write smaller test tickets across the other teams too, and track them in a separate evaluation sleeve. Now you're not betting your read of a fact sheet; you're watching live behavior across the field. When one of them starts pulling ahead, you can reallocate on short notice, or top up the one that's actually working - instead of discovering six months late that you backed the wrong horse and the right one already filled its capacity with someone else.

Which brings me to the other side of the same coin: concentration.

I see allocators develop a heavy bias toward one team - fall in love with the story, the founder, the early numbers - and overextend a large bet on them while basically ignoring the rest of the field.

First, you're not diversified. Second, you've handed a single decision the power to define your entire portfolio's outcome.

Picture it. You put the big slug on one team, eggs in one basket - and the team underperforms, or worse, slides into a real drawdown. Now you're carrying two losses at once: the obvious one, and the invisible one - all the genuinely good teams you didn't back that deserved a slice of that capital. You're sitting below the high-water mark on your largest position, and pivoting out is its own trap, because whoever you reallocate to is going to hit their own rough patch eventually too. So you freeze. Underwater and stuck.

There's no clean formula that solves this. But the principle is simple, even if the discipline is hard:

Bias is a strong word, and the whole job is managing it - managing risk, managing emotion, and never getting married to a single trade inside the framework you've been given.

The teams that survive manage the clock. The allocators that survive manage the coin. Almost everyone in this market is fighting the same two enemies - time and variance - and the ones still standing in a year are simply the ones who refused to mistake either for a relationship.

That's the read for this week.

If you know an exceptional trading team - or an allocator who should be seeing a wider, better-vetted field than the one they're choosing from - plesae do not hesistate to connect us. We reward you, properly, for any introduction that turns into business. Reach me directly at [email protected].

Stay patient. Stay honest. Stay alive.

See you next week.

Quants.Space is an institutional discovery engine for systematic and discretionary trading strategies - 130+ independent, world-class quantitative and discretionary trading teams, each with vetted track records and unique alpha sources, plus a dedicated Emerging Managers sector for early-stage teams. Our mission is simple: connect institutional capital and allocators directly with best-in-class teams, all within a secure Separately Managed Account (SMA) framework. If you're an allocator active in the SMA space, or a team opening SMA capacity for institutional tickets, get in touch at [email protected].