The second job you didn't sign up for

You launched a fund and accidentally became the COO and the IT department. An hour and a quarter with Søren Hæstrup on the business you actually started, and why most founders break there.

A couple of weeks ago I got on camera for an hour and a quarter, which for me is a small act of violence. I love writing. Video makes me want to crawl under the desk, and I said as much out loud on the recording. I did it anyway, because the conversation turned out to be one of the better ones I’ve had about what it actually takes for an emerging manager to raise early capital and survive long enough to spend it.

I sat down with Søren Hæstrup of Coherra. He ran his own hedge fund for almost 20 years and now works on, in his words, “how we can get closer to allocators and understand them better by using data rather than requesting their time.” That last clause is the interesting one. Most managers treat an allocator’s attention as something to extract. He treats it as something to spend carefully.

Søren and I go back 25 years, to my UBS days, when he was the manager raising and I was the one putting him in front of investors. Then we lost touch for the better part of two decades. We found each other again because we had both, separately, started posting on LinkedIn. He watched an interview of mine, sent me a message the same day, and here we are. Hold onto that, because it’s the argument the whole hour keeps circling back to.

I’m not going to recap the conversation. That would defeat the point. The full thing is here. What I want to do is tell you what to have in your head before you press play.

Start with the odds.

At UBS I watched a lot of managers launch, and the bank’s own arithmetic about them was blunt. You back a spread of early-stage managers knowing that one becomes a cash cow and four become dogs. That’s a revenue view rather than a survival rate, but it decides who keeps getting called back and who doesn’t. Then there’s the pattern that stings more. Talk to anyone on the sell side and they will tell you about a manager who was, for a while, the client, the one everybody wanted, and who then simply disappeared. None of that arithmetic makes it into a launch deck. It should.

My favourite client is a neon yellow elephant riding a bike in a pink tutu.

That is the most honest way I can describe the hedge fund manager I do my best work with. Around $50M under management, so roughly at breakeven. Two years of live track. A Sharpe north of one and a half, usually closer to two. Capacity-constrained, so they cap out somewhere around $300 to $500M and know it. And based nowhere near London or New York. On the call I explain why that last detail carries most of the weight. It tells you what they’re optimising for, and it isn’t only AUM.

The reason I can describe them in that much detail is that I sat down and decided who they were. Most managers never decide who their buyer is, which is why the pipeline fills up with whoever happened to turn up.

Running a hedge fund is two jobs. You only trained for one.

Søren put it plainly: “it’s two different things to run a business and run a portfolio.” The second is the one nobody warned you about. You become a jack of all trades overnight.

Look at what you’re walking away from. Inside a big firm your computer dies on a Monday, you press a button, and IT is at your desk ten minutes later. You don’t book your own flights or file your own expenses. There is a department whose job is to make your slides look good. The name on your business card opens doors before you’ve said a word. Then you leave, and every bit of that machine is gone. Who runs payroll. Who handles compliance.

The resentment does more damage than the distraction. A portfolio manager buried in work they hate starts resenting showing up every day, and that leaks into how they show up to everything else. So separate the person who runs the money from the person who runs the business. Of all the mistakes we covered this was my first, and it is the one that quietly closes funds that were never in trouble on the trading side. How you do it when you can barely afford one hire is on the call.

A name on the door does not make a business. Clients make a business.

Get comfortable selling or accept that no capital is coming. There’s no version where you sit quietly and get found. Not any more.

This is where Søren said the thing I’ve been chewing on since:

“Today you will see more successful firms with not as good a track record than you will see successful firms with a good track record if they are not visible. It’s simply so competitive, and the demand for being found and do also your investor relation obligations has changed a lot.”

Søren’s business is video distribution for the industry, so he profits when managers decide to become visible. Mine profits when they get in front of allocators. Discount us both accordingly. I’d still take the observation seriously, because he ran money for decades before he sold anything to anyone, and because I watch it play out from where I sit every week.

Returns are no longer the whole product, and the bar has moved, part of it to whether anyone can find you at all. For most portfolio managers this is the hardest ask of the lot, because selling sits about as far from what they are good at as it is possible to get. Doesn’t matter. You still have to do it.

When you’re in a drawdown, the instinct is to go quiet. Do the opposite.

You will have one. Everybody does. And the moment you least want to communicate is the moment it matters most. Over-communicate through the bad months, because the question a prospective investor is really trying to answer is who you are when you’re losing money. And whatever you do, don’t stop reporting to the manager databases because the numbers are ugly. Everyone knows that trick. A gap in your reporting says drawdown louder than the drawdown would have.

Whatever you think the raise will take, double it.

This one was Søren’s. In his day it was 18 months from first engagement to institutional allocation. Today, he reckons, it’s easily 24. My answer was cruder: whatever you think it is, double it, and build your runway for that number rather than the one you’re hoping for.

Then he put his finger on the part that actually breaks people:

“The problem often is that you get a lot of positive responses during those, let’s say, 24 months. So you feel like, oh, we are almost there. So you have so many disappointments along the way, and you just have to know that that’s completely normal. But it’s hard not to get excited about a positive response and imagining, okay, there’s going to be an allocator meeting in three weeks. We should know by then. And you don’t.”

If you tell me you want to raise in three months, I’m the wrong person and it’s a short conversation.

The meeting is the easy part. The follow-up is where people lose it.

You finally get in front of the right investor at the right time. The meeting goes well. They ask for your DDQ and your return attribution, and you walk out already knowing it will take three or four weeks to pull together. By the time you send it, the deal is gone and nobody has told you. You control the speed and the quality of your response even on the days you can’t control the numbers, and in a market where everyone is fighting for attention, that’s the part that separates people. Build the pack before you take the meeting. Most of it is standardised anyway.

Someone asked which counts for more, a backtest or a live track.

I have never seen a bad backtest. Draw your own conclusions, then go and hear what allocators actually do with the gap between the two, and where your own record really sits on their scale.

Treat every investor meeting as a 20-year relationship, including the ones that go nowhere.

People move jobs and firms. The allocator who can’t touch you today may be somewhere that can in a year’s time. And a flat NO can still be a conduit, because the person who passes on you might walk out of the room thinking someone else should meet you. So the question worth asking yourself is what they’ll say about you once you’ve left, which is a different question from whether they can tell you apart from the last five managers they saw. Be worth talking about, then stay in touch. Staying in touch when you have nothing to sell is the hard bit, and there’s one question you can ask before you leave the room that makes it far easier. I give the exact wording in the conversation.

Managers are very analytic about the markets. Less so about everything else.

Søren’s diagnosis, and he’s entitled to make it because he was one of us:

“Often hedge fund managers tend to be very analytic about the markets, and less so about a lot of other things.”

That is the whole case for writing things down. On the call I go through the second-brain system behind the near-600 investor and manager meetings I’ve documented in the last fifteen months. Every meeting transcribed, every conversation searchable, so that “I don’t remember” turns into “that allocator clears through X and passed in March on liquidity.” You automated the portfolio years ago. The raise is probably still being run on memory.

The investors you want are watching in silence.

You write, nobody likes it, and you feel like you’re talking to a wall. Then the email arrives. It’s happening to me right now, every week, several investors and tens of managers, almost all of them people who never once left a trace. By the time they book the call they’ve read enough to have qualified themselves. That costs you nothing but time. I wrote about the lurkers here, and the webinar is another hour of the same evidence.

Which brings me back to where I started. I hate video. Søren and I found each other again after twenty years because we were both, separately, doing the uncomfortable thing in public. That is our entire mechanism.

The full conversation with Søren is here. Then do the two free things I left them with. Be visible, and record everything you do as IP. Both are free, and neither requires anyone’s permission.

We ran fifteen minutes over and still never got to AI, which Søren says is an hour on its own. Another time.

Reply and tell me which of these you already know you’re getting wrong. I read every one.

Cláudia