Foreign warnings alerts, and when things get really serious, our phones start ringing even if they're in do not disturb mode. We have a cascade of alerting systems to notify us of any issues that we have with any systems. With exchanges going down because exchanges can have outages, they can stop broadcasting correct positions, all sorts of things can happen and we have to be informed of that. So the way that we sleep soundly is we build very robust risk management systems.
Imagine spending an hour with the world's greatest traders. Imagine learning from their experiences, their successes and their failures. Imagine no more. Welcome to Top Traders Unplugged, the place where you can learn from the best hedge fund managers in the world so you can take your manager due diligence or investment career to the next level.
Before we begin before we begin today's conversation, remember to keep two things in all the discussion we'll have about investment performance is about the past, and past performance does not guarantee or even infer anything about future performance. Also, understand that there's a significant risk of financial loss with all investment strategies, and you need to request and understand the specific risks from the investment manager about their product before you make investment decisions.
Here's your host, veteran hedge fund manager Nils Kostrup Larson. Welcome to another episode in the Open. Interest series on Top Traders Unplugged, hosted by Moritz Siebert. In life as well as in trading, maintaining a spirit of curiosity and open mindedness is key. And this is precisely what the Open. Interest series is all about. Join Moritz as he engages in candid.
Conversations with seasoned professionals from around the globe to uncover their insights, successes and failures, offering you a unique perspective on the investment landscape. So with no further ado, please enjoy the conversation. Hi podcast listeners, it's Moritz here and today I'm back with episode number 14 of the Open Interest series on Top Traders Unplugged. Let me wish you a happy New Year. So today it'll be a different episode because we'll be changing the asset class.
Instead of focusing on equities, commodities, currencies and bonds, we'll be talking about systematic trading in the crypto space. Yep, that's Bitcoin, Solana, Dogecoin and so on. And I'm really excited to have the founders of one of the largest and oldest crypto hedge funds on today, namely Anatoly Krakilov, the CEO, and Michael hall, the CIO of Nickel Digital in London.
We'll use the next hour or so to speak about why Anatoly and Michael decided to quit the traditional markets and start a crypto hedge fund. And also about how their business has changed in recent years, transforming from an arbitrage focused setup into a multi manager, multi strategy setup today.
Although Nickel runs a successful business with more than 250 million in assets today, it wouldn't be a complete discussion if we didn't touch on some of the nerve wracking trading days around the time of the FTX default In November of 2022, the collapse of the Luna Stablecoin in May of that same year, and how their business has coped with these events and become stronger and more resilient since then.
But before we start, let me provide you with some background on Anatalie and Michael both started Nickel Digital in 2019 and prior to that Nathalie was working at Goldman Sachs Asset Management as well as JP Morgan Asset Management in London. Michael's background meanwhile is centered around arbitrage and prop trading for hedge funds and banks, including several years at the fixed income arbitrage group of Bankers Trust in New York.
I'll stop it here so it doesn't become a monologue and instead say welcome to both of you and Natalie and Michael, thank you for coming on Open Interest. Thank you very much for having us. Absolutely, my pleasure. So let's start with an introduction to your business. How has it started? Why did you make that move away from the traditional markets or around 2019, why the decision to start a crypto hedge fund? I can give you quickly my story.
As you mentioned I was with GS at the time and in 2017 many GS clients were coming back with the question how shall we think structurally about this emerging, if you can call it an asset class and whether there is a place of these assets in portfolio construction. And that's for the first time I started to look deeper in this space.
First actually with some degree of skepticism, but then once you deep dive and look in the beautiful beauty of the blockchain architecture, it certainly captivated me. I went back to Oxford for a course, dedicated course in blockchain which became kind of point of no return. And then I started to think actively of building a hedge fund around digital assets.
And I was lucky to meet to meet Michael, who was also in London in the other part of the city thinking about the same stuff, building a hedge fund around digital assets. And with the third partner, Alec, we partnered to bring this solution to our clients. And the logic was pretty straightforward. Crypto per se is a very highly volatile asset, right? Running at volume of 60% and we thought that not everyone would be able to underwrite full volatility of 60.
But if you can lower volume to a single digits, it fundamentally fits many portfolios because it's a very alpha rich environment and lends itself to classical strategies like start arp, relative value basis funding, ARP offering much higher yields than you can achieve in equity, fixed income or commodities. Michael?
Yes, so my background is in fixed income arbitrage and some of the crypto trades one does in a crypto market neutral fund are very similar to those trades, except crypto is a much more, much less efficient market than the traditional fixed income markets.
What initially attracted me to crypto was the fact that it's a new form of money and it's a programmable form of money and we have the Internet where we have access to information on a real time basis, but when it comes to dealing with money, we don't have access to it on a real time basis. At least we didn't before crypto came along. And that was the attraction of crypto and the inefficiencies of the market for the type of trading that we could do.
And I was running managed accounts for some high net worth investors before starting nickel. And what made nickel possible was really the move away from self custody which people were doing prior to 2019. And in 2019, businesses like Copper enabled solutions like the Walled Garden, which made it possible for us to do arbitrage across multiple exchanges, but not have to do self custody. So it was a much more institutional solution.
And Nikhil has really set ourselves up to be an institutional asset manager in the largely unregulated crypto space. And we can dive into this in more detail if you'd like to. Yeah, and we will and I absolutely like to. So by the way, full disclosure, Natalie, you and I, I think we know each other since five years on and off we've been speaking about the crypto markets and I remember that you started the business around an arbitrage fund which you managed yourselves.
We will get into the multi PM, multi strat business in a few minutes, but maybe take us back to the origin. You've discovered these inefficiencies which you've just described, Michael. They are maybe we should say were, I'm not sure to what extent they are decaying, but they, you just mentioned they were larger or the markets were much more inefficient in the crypto space than for instance in, you know, traditional fixed income markets.
So that is why I reckon you had the idea of setting up this arbitrage focused fund doing basis Trades, yield based trades, all that type of stuff take us a little bit back down memory lane. Like what type of trades did you do, how did they work, how successful were they and what of that stuff is no longer working today and how have these markets changed?
So if we go back to a long way back to the fixed income basis trades, people regularly, and I used to when I was a fixed income trader, do trades between futures like the 10 Year Note Futures futures and the underlying basket of the cheapest to deliver bond into those futures. And one would be very excited to get a 25 basis point or a 50 basis point opportunity over three months. That one would then lever 50 times to get a decent return on this.
Going back in crypto in 2019, you were seeing unlevered basis opportunities. That's the difference between the price of Bitcoin and the annualized future to say the next three months deliverable future in crypto, the Bitcoin future of 40% unlevered. So one could get those type of returns without using any leverage. Now one did take some exchange risk because one had to move bitcoin assets to an exchange like Bitmex, which was the main Exchange back in 2019.
And there were various hedging risks associated with that. It's not a risk free trade. For example, if you've gone long bitcoin, short bitcoin futures and put some bitcoin on Bitmex to collateralize your short bitcoin position, if there's a very big rally in Bitcoin, you need to be very quick to move money onto the exchange, not to be liquidated.
And I'm sure your listeners are familiar with the memes of Arthur Hayes and the crypto chainsaw and the massive liquidations that used to take place in the volatility of the markets. But that was the. The basis trade is a good example. Now how has that changed as more people have come into the market? You can now go to venerable institutions like Goldman Sachs and they will give you a package of Bitcoin versus the CME futures basis and offered that to investors.
And that's led more capital coming in to that inefficient market which has driven the spread tighter. Now it's not to say the trade has gone away. It means the trade normally is not at an attractive level. From time to time when the market gets very leveraged, gets very frothy, more interest comes in and the basis trade widens. But the basis trade can't be your sole reason for running a strategy in crypto market. Neutral, one has to have other strategies around.
So again, going back to that time, there were other strategies. Like you could lend money to traders on bitfinex and get 40% yields. You could do triangular arbitrage between different exchanges and extract funds. That way there were funding opportunities. So there were still many crypto opportunities and to an extent a lot of them still exist. Yes, and by the way, we participated in some of these as well, especially the cash and carry trade.
And I remember when FTX became a thing and they became really popular, some of that basis went away because you could deposit all sorts of collateral on FTX and kind of like do the trade there. Now I haven't done that trade in a long, long time. But here's a question, I didn't actually plan that. Now with the emergence of the Bitcoin ETFs in the States and you know you still have the CME futures contract.
A couple of months ago I had a look and I was surprised to see the basis still at kind of like 12% or 11% annualized. Maybe it was only like for a blip or for that day, but I was surprised to see it's still that high because I was thinking, well, now that you have essentially a, you know, an efficient tracking product of spot Bitcoin and you have this futures contract, shouldn't that basis collapse? Question mark, I'd like to ask you, why is it still so wide?
Because the counterparty risk is now essentially gone. Yes, but I think what happens is managers hit their capacity on the trade, so the people who are doing the trade have a limit of what they can put in and they get their full size on. Yet the demand for leverage, the demand for futures still increase, is still there and it pushes the basis higher. Supply, demand, it's supply demand and it's people hitting their capacity limits.
And then of course the opportunity goes away and all that capacity then is just effectively disappears, or rather the trading money looking for it disappears. One thing we saw after ftx, for example, is FTX did squeeze the basis much tighter. But after ftx, many hedge funds shut down their crypto businesses. That led to capital flowing out of the crypto arbitrage space and that meant the basis was more volatile and became attractive again. So it's this ebb and flow of capital into the markets.
There's demand for leverage and there's also which then capital flows in to take care of and then the capital exits because the opportunity goes. So it's very much a, an ebbing and flowing market, but it hasn't disappeared. And actually already see last year upon the launch of Bitcoin etf, between January and March, yield on basis trades spiked to over 30%. It was 30 to 40% annualized. Of course it's not very sustainable. It's collapsed immediately as market went in the sideway after March.
But yeah, it's still happening, right? There are still this kind of dislocations in the market. They do exist, but exactly because of this trade collapsing so quickly. As Michael mentioned, it's important to have a range of strategy in your portfolio rather than relying on basis as your core strategy because it's not very sustainable, very sensitive to the rally in the market and cannot really cover 12 months a year. It's maybe just spikes right throughout the cycle.
Which is why this wasn't the only trade that you ran when you had the arbitrage fund. I guess you had, you know, perpetual futures versus fixed term futures and also curvature trades, you know, stuff like that. Now I presume the alpha in these trades has decayed a bit. Like I'm just observing it as an outsider. There's probably more participants, more smart money in these markets these days and it's become more difficult.
So maybe Anatalie, tell us the evolution or the progression of your business where you decided, okay, we're stopping the arbitrage fund or we're trading ourselves and we're migrating into a multi PM multi strategy setup where we're using the brains of a number of independent people and pods to trade in these markets and extract a different source of alpha. Absolutely.
So our logic back in 2020 when we already had arbitrage fund for over a year, was that actually we're coming across incredible talent in the space, people running a much wider range of strategies than we could have run of being a single manager fund. Specifically start, AR relative value trades, all quants natural. Because we don't do, by the way, you should have mentioned, we don't do discretionary trading. All strategies we run are pure quant.
And we saw emerging range of managers with whom we were either competing or they had strategy which actually could have been complementary to our strategy. And that led to the idea of building a multi PM fund whereby we are kind of aggregating different strategies under one roof, adding start, arp, NRV as natural expansion of our ammunition, if you will. And that led to the idea of diversified alpha which at some stage coexisted with arbitrage.
But again, if you look on multi PM structure versus single manager fund, multi PM provide by design better and More stable return. And Millennium would be a great reference point here. Right. So Millennium has been in existence since 1989 with one negative year and 34 positive years and that's certainly kind of a good reference point. So from that perspective we decided to build a multi PM fund and today we have 36 different fundamentals trading teams I.e. pots under one umbrella.
And there is a great diversification across strategies including as we mentioned, start ar, RV basis funding AR mean reversion, short term trend following, money market making and that diversification actually creates stability of returns. So at some stage we decided that we're going to wind down arbitrage, essentially merging it into the diversified alpha to avoid having two alpha products within nickel famido funds and avoid this unnecessary competition.
Yeah, so basically now we have one single alpha implementation being diversified alpha. And as a byproduct you've also changed your careers inside Nickel Digital from being traders to now more probably running operations risk management oversight of these pots. It's a different role that you now need to fulfill. Right, so our core focus perhaps here is risk management. It's a manager, selection manager testing, validation, incubation and ultimately risk managing these managers individual pots.
So to that end, given that crypto runs 24,7, not only your underlying strategies have to be fully systematized, but risk management has to be fully systematic as well. We tested a couple of systems off the shelf solutions back five years ago and concluded the granularity of risk oversight wasn't up to our liking. Ultimately we decided to build both RMS and EMS in house.
We have 30 people in the firm and roughly half of us are software engineers who actually contributed to building these very powerful systems. So yeah, once managers are onboarded, there is a gradual process as to how we provide testing capital before scaling them up to the full size.
But in every single part of this journey we have perfect visibility in the underlying portfolio and running real time observation of the book against risk metrics which has been pre agreed with each individual PM before first dollar is awarded. So to some extent, yes, we are now in business of manager selection and risk managing our portfolio.
So Michael Anatalie has just mentioned 36 pots and he also mentioned 24, 7, 365 trading, which for you guys means there's no Saturday off and there's also no Sunday off. So you have a fully equipped office seven days a week. So what crypto risk management forces one to do is it forces one to automate. So you can do one of two things. You either have a round the clock Round the globe operation managing risk.
And that doesn't always work because we've heard of situations that the guy in Singapore was asleep when the alarms were going off. So that's one option. The option number two is you automate everything and we have risk management systems that have warnings, alerts and when things get really serious our phones start ringing even if they're in do not disturb mode.
We have a cascade of alerting systems to notify us of any issues that we have with any systems with exchanges going down because exchanges can have outages, they can stop broadcasting, correct positions, all sorts of things can happen and we have to be informed of that. So the way that we sleep soundly is we build very robust risk management systems with pre set risk limits and controls and we know we're going to get alerted now that happens from time to time.
So when we saw the USDC DPEG for example, everything started going off because clearly that influences and has an impact on the balances that our managers have at exchanges and we started getting notified of that as soon as it happened. And that's good from two reasons.
One, we can control risk and two, we can see potential trading opportunities and we can highlight those to managers who should be taking advantage of these just to make sure they are and we can deploy more capital to them to enable to take advantage of these opportunities as we see them happen in real time. This is not a fund of funds where we do not know what the manager's positions are.
We know our managers positions on a real time basis and we work with them to give them capital to help them exploit all the opportunities that they have. Yeah, you have oversight and control over them is my understanding, because they're trading into a central book. But I want to get down that rabbit hole in a few minutes there. There's more to unpack there. But when we stay with the 36 parts that you have onboarded presently, where are they based? Are they all around the world?
Do you ask them to come to your London office? Can they be wherever they want to be? Do you need to meet them in person during your due diligence process? How does, how does it work and how do you find them? Ah, those are lots of questions. I firstly want to pick up on something you said about central book. I want to be very clear that we do not run a centre book.
There's no central risk book, there's no internal alpha capture, we don't run any overlays, we allocate capital to managers, we don't copy trade Managers. In fact, in all the contracts with our managers, we have a specific section covering intellectual property. And in that we commit not to reverse engineer their strategies or copy trade them or otherwise take advantage of their positions. We want to find the top talent in the world and the top talent in the world doesn't want to be copy traded.
So we absolutely don't do this. So hence the central risk book. If you had a central risk book, it doesn't necessarily mean that you would copy trade your PMs. You could detect something like, oh, everybody has a massively long eth position or Solana position on and it kind of like exceeds the risk budget that you would allocate to Solana inside of the diversified portfolio. And therefore you would say no as a factor, I'm reducing Solana exposure. You're not doing that. We're seeing this anyway.
We have this full visibility. Even if how it's central book. But, but the problem with this, and I've seen this when I worked at Bankers Trust and you had, whenever you have a shared book, it's my gain and somebody else's loss. So imagine you say, okay, we have too much Solana risk across the portfolio and somebody in the inside nickel decides to go short Solana and Solana goes up. Does that come off their bonus?
Like we have to pay performance fees to our managers and if it goes the other way and the manager makes money, does the manager get paid a bonus for that? Like, it's a very bad rabbit hole to go down. And the way we don't go down the rabbit hole is we don't run overlays. If we think there's a concentration in any one factor, we will potentially reduce an allocation to the POD manager who has a big exposure to that factor.
We would go to them and say, listen, you and several other pods have a large exposure. We would like you to cut this exposure. And in the same way as we thought people were too exposed to basis or funding or any particular type of trade, this is how we would deal with that risk issue. We wouldn't try and hedge it ourselves. Right.
You're essentially getting to the same point by dynamically changing their notional trade allocation and, and maybe quite rapidly sometimes so that you're not overexposed to a certain factor or position. It's does not. What you've just described is interesting about the compensation because it doesn't at that point address the net debt performance fee. You could still have, you know, out of your 36PMs, you could have 30 that make money and sticks that lose money, which is probably expected.
Right? But as Michael was saying, if it's my gain and somebody else's loss, the people that make the money still want to be paid the bonus. But do you net everybody's performance fee or does nickel digital as a business come up with a delta and pay everyone accordingly? Everybody gets paid according to their contract. We have performance formulaic performance fees in all of our contracts with managers and everybody gets paid on a quarterly basis what their formula is.
Last year we paid out nineteen and a half million dollars in performance fees. Now that may seem outrageous for a fund, but our allocators made a very good return last year, which they're very happy with. So there is. But netting is a very big issue in multi manager funds. We noticed that everybody talks about it but nobody writes about it. So last year we decided to write a paper and we published a paper late last year which we.
I don't know if you saw this on LinkedIn, but we published a paper specifically on netting risk and how we look to reduce netting risk between the managers. Because it is a serious risk and it's something that you can never get rid of it, but you can minimize it. Well, actually you can get rid of it. You can put in invest in one manager and if you have one manager, you have no netting.
But of course then you wouldn't be a multi manager fund if you had one manager and you lose all the diversification benefits of it. So with diversification benefits comes a netting cost. So this whole idea of diversification is a free lunch. It's not a free lunch in multi manager funds. It has a cost, netting cost. But there are ways of minimizing that netting cost. And in that paper we delve into the ways that we look to do. That partially by controlling the downside risk. Right.
And that's where your limits come to play. Max drawdown and net gross, which we control very tightly. Are you guys passing through costs like some of the multi PM multi strat shops do, or do you charge a fixed management fee on the fund and you essentially cover the remainder of all costs? So the answer is it's a 2 and 20 standard hedge fund fees with no pass pass through per se. There is a netting cost, as Michael mentioned, which naturally is absorbed by the underlying investors.
However, there are no hard costs which are usually passed within multi manager structures, traditional ones. Right. And in the usual structure you would have PMs moving from one hedge fund to another and usually there is a significant upfront cost being sign up bonuses and all this stuff which usually just kind of passed through to the underlying investors.
In our case we do not have these costs and perhaps it's worth spending a couple of minutes as to design of the fund because you absolutely correctly mentioned we are not fund of funds, we are a multi pm. And the core difference is whilst both fund of funds and multi PM fund would rely on a kind of different talents, right kind of different trading teams, fund of funds would invest in other funds and by virtue of that lose control over capital at least for a month, realistically much longer.
And funder funds do not have a luxury of visibility, perfect visibility of the underlying positions. Multi manager funds on the other hand have perfect visibility, control of underlying capital and can withdraw capital from any given PM at any time. And that's exactly model we are following. So we're never writing checks to any third party funds or managers per se. Now managers by design are all external.
And the logic here is when you have managers sitting on your desk, there is a legacy reason for that. And one of the legacy reason, if you look back in 90s when many multi manager funds emerged, the Internet wasn't there. So but the very least you wanted your manager to your PM to be based on the desk next door to you so that you can come tap on the shoulder saying like show me your risk positions. Now where we are today in 20, 24, 25, we don't need to talk to anyone's shoulder.
We see risk as it happens, real time, irrespective where a manager sits next on the next desk and the next building or in Singapore for that matter. Because capital stays with us and we control the capital, managers can instruct and send signals pretty much from anywhere in the world. So that answer your question. Managers are actually located across multiple countries, multiple jurisdictions, but capital is controlled here in London.
And from our perspective, as long as we can have a centralized risk management, we're happy for managers to be located anywhere in a decentralized fashion. We're getting perfect access to the brain power without losing control over capital. Let's speak about the onboard onboarding process you mentioned. They can be really anywhere and everywhere from all countries in the world. How do you find them and how do you onboard them? Like due diligence, all that type of process.
So not every country in the world, North Korea, Iran, there are some countries on that list that we would not. You wouldn't like that managers should not be located there. But coming on to what Anatoly mentioned about managers working remotely, it comes down to the incentive compatibility of the setup of the fund. Everywhere where we've looked at the fund and the economics of the fund, we've said how can we make this as incentive compatible as possible?
And that means that we pay higher performance fees to managers, but we don't pay signing on bonuses to managers like a large multi manager would do. We don't pay management fees either because all the things that a POD manager would spend the management manager fee on, we provide them with that. So a manager coming onto the platform doesn't have to set up a fund in BVI or Cayman. They don't have to hire COO or compliance officers.
They don't have to hire investor relations to look for money because we provide them with money. And they don't have to write any newsletters. We write a newsletter for our investors, but they don't have to do all of these overhead tasks that a manager typically would. Setting up a fund to fund. What this means is having the managers remote means that we really cut the costs down for our investors.
So the only they pay a 2% management fee that we spend on compliance and investor relations and raising the money, et cetera. We have some funding fixed fund expenses like audit fees and administration fees and directors fees, but they're pretty very low as a level of the AUM of the fund. And everything else is on a performance fee basis. If there's no performance, there's no fee and it makes it very efficient for the managers. Now coming on how we find the managers.
Hopefully there'll be some managers listening to this podcast who'll think, well this sounds really interesting, I should be signing up to trade for nickel. And why would a manager want to do that? What's the attraction of them coming onto the platform? Well, the first thing is high performance fees paid quarterly. So we pay performance fees of up to 50%. Now that sounds like a lot, but to get that 50% performance fee, the manager has to be making a 50% return with a very high Sortino ratio.
But we think that's fair if it's it's all about the net returns to our allocators. The other reason why they would come to us, It's a contractual B2B payout so that it's not a discretionary bonus. And at the end of the year we're telling them, well, I'm sorry, the other pods didn't make money so we can't pay you. That doesn't happen. We have contractual payouts with each of the pods and if they've made money they get paid no matter what happens to the rest of the book.
We've got very good fee tiers on the exchanges because we agglomerate all of the volume of all the managers trading. And the new manager coming onto the platform who's managing some test allocation for us gets that fee benefit of the big managers who are trading like the $50,100,000,000 allocations. And it's worth mentioning having traded $86,000,000,000 last 12 months, certainly we're in a best feat years with every single exchange.
And even for managers who trade relatively kind of mid frequency trading trades or kind of strategies, they still benefit from the fee tiers which we have already on the fund level you asked like. Why demand had come to nickel. No copy trading or reverse engineering. This is a big deal. Like I'm stunned when I hear that groups big multi managers run central risk books or they call it internal alpha capture. These great euphemisms for meaning.
We will just rip off the IP of the pods trading for us. I saw this myself. I used to trade at a $14 billion fund. I had a $200 million allocation to trade fixed income relative value. And every time I did a trade that the major fund could trade, they copied it. And I didn't get paid any performance fee for that. But that was my intellectual property that was effectively taken from me.
And if we want to attract the best traders and the only way we make the returns that we do is because we have a fantastic stable of traders trading for us. We don't rip them off, we don't steal their intellectual property. That's how we get the best traders. We want to create this selection bias where there's no management fee, there's no fee to walk in the door, but you do get paid a very high performance fee and you get all the capital you need to take the capacity in your trade.
We don't say, well we'll give you half the capacity and we'll just copy trade you and take the rest of the capacity in your trade. We don't do that. If there's capacity in the trade, the manager gets all of that and that's all for them. The managers have full independence to do what they want to do to be anywhere, well, except for those five countries I mentioned. And we don't ask for exclusivity.
And we are, we don't want, we don't ask for exclusivity because it's the same way that West Berlin didn't have any walls around it when East Berlin did have walls around it. You don't need to stop people leaving West Berlin to rush into East Berlin or you didn't need to do that. Whereas we don't feel we have to lock people in with restrictive covenants, long gardening leaves, sit outs, all of this stuff. We think that managers should want to trade for us because we give them the best conditions.
So the way we create that for ourselves is we do that by giving them the best trading conditions. So it's again as I've mentioned, high performance fees, not copy trading them. They own their ip. All of these reasons are reasons why managers should come and trade with us. Again. We're not forcing anyone to trade with us. Slavery is illegal. But we want people to come and trade for us. And we've got managers who've been with us for a very long period of time.
We have very good stable relationships with them, which means we can work with them through drawdowns. When they have drawdowns, we have a manager who's made us money. We will agree with them together that maybe we reduce the capital while they figure out what the issuers with the system, they fix the system and then we scale them back up again.
So we've had managers who've been on, on the platform now for four years, pretty much since we've started and we have very good, very deep relationships with them. Really interesting, Michael. I think from what I can take away the two things that you've just mentioned, the euphemism of the internal alpha capture, which is a lovely three word piece, you're not, you're not doing that. That is a differentiating factor between you and other multipam pot shops.
And then also the non exclusivity because that is, I've never heard this like usually when you run a pot at another firm that is where you trade and you're kind of like you're stuck there and when you leave you're on a non compete for not just a month but maybe a year or two or something like that. Right. The fact that they're not exclusive go into that again, does that mean that they can prop trade on the side their private account?
Could they also trade for you and at the same time sign up for Citadel or Millennium or say for Brevan Howard because Brevin Howard has a digital assets business. Right. And trade there as well. That's what not exclusive means. They're free to do that, but we would rather that they traded for us. But they are free to trade for other people. And the reason we say that, and we can say that with confidence, is that we think they will make more money trading for us than they will for the others.
Because the terms that you offer are superior to these terms that the other firms would offer. Exactly. That's what they tell us. That's what the pods tell us. Okay. And because there's limited capacity in everything in the world, they should really be allocating their trading capacity to Nickel's operations as opposed to somebody else's, which is this logical decision. This is what we find. Okay. And the beauty of our model is that there is no fixed pay.
As Michael mentioned, it can go up to 50 and that's what appeals to the very powerful and confident traders. Right. But again, when unwilling to lock ourself in a given number, but rather we offer what is called metrics, sorry, Sortino metrics, whereby you have absolute return on one, say X axis and then Sortino ratio on Y axis and the intersection of two gives you exactly perfect visibility how much you're going to be paid and the better you perform, the higher your payout.
And there is never dispute, we've been in this business for four years, we never had a dispute with single port as to how much they're going to be paid over the last quarter. They don't have to be nice to us, they don't have to be politically correct. Show me the money. Right. Kind of. That's what is perfect, transparent transparency. And we see managers opting for that formula rather than a discretionary unclear structures. Yeah, fair enough. I, I, I like that.
Let's stick with the numbers you've just mentioned. Cino ratio and maybe we, not everybody knows what a Zartino ratio is. Let's use the devilish Sharp ratio which probably everybody gets their head around these days. If you compare that, just to put it into perspective, like the cross section of your pots, what is the average Sharpe ratio that you see in the crypto space when they trade? So it varies from section to section.
So clearly people like people doing funding arbitrage have extremely high Sharpe ratios because they never lose money. The volatility of their returns is very low. Having said that, one has tail risks which around the exchanges, for example, so there are tail risks on certain pods.
If you go to the other end and you look at short term directional bots, which are trend followers, but trend followers who will go from long to short maybe four or five times in a day, they have lower Sharpe ratios, could be around one and a half to two, but they still provide value to a diversified portfolio.
And Our skill, our aim is to combine market makers, cross product arbitrage, stat arb, short term directional basis trading, all of these different strategies into one set of returns for investors. That's attractive and so far we've been able to do that and that is our goal.
So a 1.5 to 2.5 Sharpe ratio for short term directional trend following is quite nice when you compare that to what you can get in the traditional markets where I would venture a guess that you're not getting 1.5 to 2.5 on short term intraday directional trend following unless you are super duper good. But by the way, this is one of the sub strategies within the fund. The fund itself runs on a much higher Sharpe ratio because we access multiple strategies, not only short term trend following.
Yeah. So testament to the fact that these markets that you operate in and your PMs operate in offer more and more valuable opportunity sets to trade and capture alpha than many of the traditional markets. So I completely understand why you've built a business around. It makes economic sense to say the very least. I mentioned earlier the platform, I think Anatil, you mentioned the tiers, like the fee tiers on exchanges that you're using. Are you using one exchange?
Do you have a favorite destination, Binance for instance, or are you kind of set up with multiple exchanges and then Copper as a custodian? Give us a little bit of the backbone, the tech backbone and how you're connected. So we look for the best providers in each of those areas, exchanges and custody. And we're really driven by where our underlying portfolio managers want to trade.
Typically they've been running a trading strategy on one of the exchanges and they come to us and they say, you know, this strategy runs on okx, the strategy runs on Binance or Bybit. We have access to all those exchanges we have because we trade, we can agglomerate the volumes across the exchanges. We are on very good fee tiers for those. 90% of our volume is maker volume and that gives us.
We're often getting rebates from the exchanges, which again makes it attractive to trade on those exchanges for the managers. And the managers benefit from those rebates by the way, not us. It goes back into their trading account so it reduces their trading costs. So that's another reason for managers to be trading with us. On the custody side. We're always looking for the best custody solution. After ftx we decided we would not put assets on exchanges anymore.
It was a case of like fool me once, shame on you, fool Me twice. Shame on me. And we didn't want to be fooled a second time. So everything moved to off exchange settlement and we continue to kind of push that. Right now we're looking at. We've signed on for Signum to trade with Binance with their. It's called, it's Tri Party version 2 it's known as.
And the idea is is the funds are with a Swiss regulated bank and they don't even touch the exchange and the advantage of that is we get yield so we can get a T bill yield or we can use BIDL and the ERC20 and get a yield on that and so we can move collateral around very quickly between banks. But again it's the latest custody solution. We're always looking for the best quality custodian solution for our pods and for the allocators.
I know the folks at Signum Bank, Matthias Imbach and the guys, I think they're also running a Singapore business in addition to the just the business that they have in Zurich. You've mentioned FTX which is now a bit more than two years in the rear viewer mirror. Would you say that that has been. Or this year I presume with you know Luna and then FTX later on in November has that's been. Is probably has been the most stressful period for your business. Is. Is that fair to say for sure.
And of course the counterparty risk was very high on our agenda ever since we launched nickel FTX wasn't completely unexpected event right in the industry. Having said that we had counterparty risk limits across all exchanges. We traded in the preftx world which was roughly 20% for larger exchanges and as small as 2% for the second year. Small exchanges FTX was the second largest derivative venue at the time which naturally would qualify for the larger exposure.
We limited still it to 10% rather than 20. When FTX happened we had our capital sitting on the exchange within this kind of nine and a half percent at a time. And whilst we had a clear understanding this is recoverable at some stage it was a fair approach to market down to zero with the expectations recovery will follow at some stage which actually following year we recovered majority of this capital all from the exchange from ftx.
But what we've done ever since that we have withdrawn at the time of FTX from all the exchanges saying we would come back only to those exchanges who would accept off exchange settlement tripartite agreements which Michael mentioned and as it stands today we only trade on those exchanges who would comply with these requirements. It generally Captures all usual suspects. All the largest exchanges but those smaller ones who were unwilling to accept just for us non existent pretty much.
I know Natalie, you like sailing. So has everything been plain sailing post FTX or have there been some episodes that were now nerve wracking as well that caused you to lose some hair? No, I think kind of in the post FTX world it was relatively straightforward because we kept adding additional pots, kept expanding strategies. There was no distress so far in, in the market and actually we went through the few important market events which validated the strategy which we run.
For example, August 5th last year when Japanese market sold off by 13% essentially being the worst day in 35 years.
The way strategy reacted to to these events was very telling because we don't trade equities but naturally all risk assets got affected by Japanese salo you had NASDAQ minus 5% you had crypto actually selling off by 22% that was E and Bitcoin by 16% and that happened in the middle of the night because when Japan was ahead of London e Soldo by 22% at 02:00 London time, how strategy reacted?
There was not a kind of dent in the strategy because there is no market exposure per se and as a matter of fact we made money that day. But core focus was protect capital and it kind of played extremely well. So from that perspective market wasn't a plain sale but the portfolio performed extremely well Navigating this pitfalls, Michael, I'd like. To bring it back to the onboarding and due diligence process of these parts.
I think Anatalie just mentioned how things are marked and you know, something might be down 20% at 2pm or 2am London time. But relative to what? There is no close, there is no settlement. It's kind of like there's this ongoing 24, seven, you know, progression of prices. So how do you give us a little bit of background like how do you work with that?
Because it's quite different to a, say you know, the futures contract or like a 10 year note futures contract that you mentioned Michael, which you traded way back when. Yes, you had a CBOT settlement time, that was it, it's done. So we compute the underlying manager's equity on a real time basis on a fill by fill or a price change by price change basis. And we gather return statistics that we use to analyze returns every four hours so we have a volatility and a return for that four hour period.
And then that's our, our basis when we do portfolio analysis. But Essentially we are looking at the portfolio for how much it's done in the last, how much it's made in the last 24 hours, this month, this quarter and so on and so on. So we have a very good handle on the portfolio equity at any time. That's very important. That's part of our risk management process. So we're very aware of all these metrics and what we should be doing.
One other metric apart from the performance fee that we put in the contract with the managers is the peak to trough drawdown the manager can have. Now we smooth the P and L because we don't want to have a sudden performance spike. We readjust the manager's high watermark. So we tend to adjust the high water marks on a, like a daily basis. We will automatically reset high water marks.
We don't reset a high water mark on a spike because it might lead to a manager being stopped out, which we don't want to have. But we reset the high water marks. The managers know what their risk, their peak to trough drawdown risk limits are and before we reach those limits we are communicating with the managers and we're discussing with them like what's been happening and so on, just so there are no surprises. And it works very well.
We had, as Anatoly said, we've never had a dispute in the four years with a manager. We have a very collaborative, cooperative way that we work with the managers and it's, it's in our interest to help them make money and develop them because if they make money, we make money. It's that simple.
We very much have aligned interests and this is why we wanted to get rid of all these copy trading and lockouts and sit outs and stuff because that's when the interests start to diverge and then you start to get odd behaviors, politics and so on which we don't want to have. We want to have a very simple clean life where people make money and they get paid for making money.
And to your point of 24, seven, if you think traditional markets would trade six and a half hours a day, five days a week, which kind of in the context of 24 hours comes to just 19% of the time whilst of the physical time. Right. While scripted trades 100% of the time. Which actually creates this very high demand for the risk control systems and execution management systems.
So the risk management which we run internally, the system is called risk Zeus as the God and the logic is Zeus because to be as God you need to have a be omnipresent And B, be able to strike. And that's exactly what system does. It has visibility of every trade on a single field by field basis.
Actually the system collects over a hundred million data points a day from the portfolio for the purpose of subsequent risk analysis and portfolio construction and is able to strike, I. E. Cut the risk if they need be. So that's how we approach the risk management and control over the portfolio. I remember I once mentioned to a client that trading crypto 24, 7, 365 is actually risk reducing and less risky than trading in the traditional markets.
I remember they look at me a little bit strange because you know, crypto is high volatility and you have all these failures. But I mentioned that from a perspective of like you just mentioned markets being closed. You know, here you have markets which are accessible and open all the time. When you look at the traditional markets, they don't trade 24 hours a day. Something like the S and P E mini contract trades 23 hours or something like that.
But like a lot of the soft contracts, they trade a couple of hours, including the electronic sessions. And then you have the weekend and then you have public holidays, et cetera, et cetera. So for most of the year the market is actually closed and you cannot react. In crypto you can, and that is something that allows you to improve your risk management. Yes, that's absolutely true.
Because one thing you don't have in crypto, you don't have this gap risk that you have in traditional markets where somebody bombs somebody over the weekend and the oil market is closed and you get in on Sunday night and suddenly the market market's moved by you know, 20, 30% and you're stopped out of all your positions and you haven't even had a chance to trade.
And it's, that doesn't happen in crypto because okay, they can be large movements in crypto, but when we see these like you know, 10% days, they happen little by little and they're monitored as they happen and we can move to take risk off as that happens. But in a sense we don't aggressively cut managers risk. Normally the managers are cutting their risk before we've ever asked them to do so.
Although we do have the power to strike with this risk system, we don't do that because again, it's collaborative and the managers are clearly trying to preserve their trading capital because that's how they get paid. And this incentive compatible trading structure means that they're very concerned they're not getting a management fee the way they get Paid is on performance. So when their performance starts to suffer, they move very quickly to cut their risk.
We haven't had to do that in the four years we've been trading, but we have the ability to do. But we've never had to do it. We even had instances when managers would come back and return capital on their own accord rather than ask requesting this capital kind of withdrawal. The reason being, as we mentioned, Sortino ratio. Sortino would suffer if you kind of going into a period of say flat performance.
So from the manager's perspective, they would rather return this capital to preserve certain they already built over previous periods because that impacts their payout. And we would gladly accept this capital back because we can relocate to other managers without us needing to cut them. So it's very incentive compatible on multiple levels. Get it? When we look at the pods and I think you're at 36 now, you weren't at 36 probably two or three years ago, I presume there was a smaller number.
When we just look at the average turnover, how many of these pods are closed in average here in percentage terms. So we funded 97 pods and we currently have 36 active pods. And so that means that we've lost a decent number of pods. Now it comes back to the way that we run the business. There are two mistakes we make as managers. Type one errors and type two errors. A type one error is we define we give money to a manager and they lose money.
A type 2 error is we don't give money to money to a manager that we should have given money to and they make money. So our process, our onboarding process is made up of two parts. There's a screening that goes on before we fund, which looks at their strategies, dives into their risk management, the legal compliance they need to go through and KYC and world check, you know, fraud checks that we do before we give them money.
And then there's a second part that happens when we make a small allocation. And this is the beauty of crypto, because in crypto you can make 100k allocation to manager and they can trade it as if they were trading a million in the futures space. You can't trade half or tenth of a natural gas future. And so that limits the diversification a manager can have when they're trading a small amount of capital. Whereas in crypto you can trade a 10,000th of a bitcoin contract and it's easy to do that.
Now there is some question about is that repetitive as you scale a manager up and they get larger and larger, but it gives us this ability to do investment due diligence on actual manager returns. And so we fund managers initially with 100k and they're then in our risk management system. We see all the trading they're doing, we see how they manage their risk.
And typically the managers who don't make it to be fully upsized, it's not because they lose money, they just don't make as much money as they thought they would make. So as I said we're looking for the minimum Sharpe ratio is around one and a half, maybe it's less for non correlated strategy. But if a manager is producing say 5% return with a 0.5 sharp, we aren't going to upsize that manager. And the manager knows that as well. Typically the back test may have shown a much better performance.
But when they come in and we give them actual capital and we see that, we then don't scale those managers up. So that's the reason for the number 97 pods we funded with incubation capital and we now have 36 active pods on the platform. It does mean that some pods have left but it that's because they their returns weren't up to what they thought they were going to be. Perhaps their alpha decayed, the environment changed, something happened. But that's how we protect our allocators.
And it also makes sense for the pods. The pods have had a chance to test something and we often will give a podcast that starts off with 100k, just a small allocation below that just for them to test and run things while they figure things out. And once they figure things out, then we're very happy to start aggressively scaling them up. So we've scaled up manager from 100k to 16 million notional in 5 months as an example.
And we can scale very quickly for high return, high sharp, low correlation managers. And we are very happy to do that. And it's important to realize the difference between 97 and 37 is not that 60 lost money. As Michael mentioned, only few of them really would lose and hit the limits. And it's realistically perhaps single digits majority would just not be up to the standard. We're looking in terms of absolute returns.
And even if they were to lose money, usually we would spot this in the early stage when 100k allocation is made. And let's be honest, 100k is not the amount they're willing to manage neither Believe us, we're willing to waste our time but it allows us to kind of have this well informed view before we scale them up how they trade.
And given that all of them are high to mid frequency traders, within the first three months they can easily have a hundred thousand individual trades, which gives us a statistically significant data set to have kind of view whether we'll scale them up or not. But if assume they want to lose 20% of the initial allocation. Let's go off into math. 100k is 5bps of the fund. AUM if they were to lose 20% of 5 bips they would lose 1 bip. Can we afford that? Of course we can.
But that allows, as Michael mentioned, to avoid this type 2 error when we do not test enough bots. So we're willing to engage with multiple trading teams and give them opportunities to demonstrate their skill set. I'd absolutely would like to invite you guys back for a second conversation in the interest of time because I have so many more questions about bid offer spreads and liquidity and market impact.
Like we didn't talk a lot about these technicalities which are usually in enjoy like you know, types of strategies that you run. But in the interest of time we're not going to be packing this into today's episode. That would be just too much of an overkill. But maybe to to wrap it up when you look at your business bringing it back to nickel digital again, what is the if there is a like a long term vision for the fund, where do you see the biggest opportunities?
Maybe some challenges down the road. Maybe we can stop it. There's so we hate saying no to anybody and in the past we've been saying no to allocators who don't like the particular flavor of diversified alpha. Perhaps it's not volatile enough for them. They want to have a more punchy, more leveraged version. Perhaps they'd like it with a slightly different flavor of managers. Fewer trend followers, more stat arbe and we haven't been able to give that to them.
And so what we're doing this year and you'll see us start to launch, is doing customized fund mandates where we can do fund of ones for managers. And these fund of ones will have specific risk profiles that the allocators want. So if an allocator wants trend followers, they can have just trend followers and they can lever them up as well if they'd like that. So what they have access to is all of the managers within the diversified alpha, but they can also bring their own pods as well.
So we have a potential fund of one who wants to bring some of the trend followers that they currently allocate to and see how those guys do when they are in the crypto markets. And we can do this now, we wouldn't do this within Diversified Alpha. We wouldn't let somebody show up and just say, hey, put some of my pods in Diversified Alpha. But if it's their own fund, we're happy to do that.
So what you'll see from us in the next year is lots of custom mandates, which we think is great because it brings more traditional investors into the crypto space, brings more capital into the crypto space and provides more funding for these great managers we have within the diversified Alphabet business. And of course it's really triggered by existing allocators from traditional space who are looking at crypto as an emerging market. They would like to engage.
But as we just discussed, there is a very specific idiosyncratical technical requirements to engage with this market. And unless you have this infrastructure, you cannot really properly allocate capital. So by leveraging nickel infrastructure, they can build customized portfolio specifically designed to access crypto strategies. And we're in good position to provide that because we have been building this for the last five years. Excellent.
Well, kudos to you for building such a great business in the digital asset space. It is really something. And let's leave it there. Michael and Anatoly, thank you so much for joining me on the Open Interest series today.
It was great speaking with both of you and and I'm sure our listeners will find it interesting our expedition into the digital asset trading space today, which is not what we usually do, I certainly found it interesting and I also learned some new things, which is always a good thing, right? So as usual, we'll include the most important points of today's discussion in our show notes and please don't hesitate to contact us should you have any questions.
You can reach us@infooptraders unplugged.com and we're always happy to hear or read from our listeners. So thanks again for listening. And until next time on Top Traders Unplugged. Thanks for listening to Top Traders Unplugged. If you feel you learnt something of value from today's episode, the best way to stay updated is to go on over to itunes and subscribe to the show so that you'll be sure to get all the new episodes as they're released. We have some amazing guests lined up for you.
And to ensure our show continues, continues to grow, please leave us an honest rating and review in itunes. It only takes a minute and it's the best way to show us you love the podcast. We'll see you next time on Top Traders Unplugged.