Coinviction

Thesis driven portfolio management

My impression from reading my feed the last few years is that the CT hivemind could be best described as “a degenerate gambling paranoid schizophrenic”. While regularly trading 70-100 vol assets, the feed still manages to freak out on mundane -4% moves after a +50% ripper. But behind the façade of hyperbole, memes/shitposting, and LARPing as a cat, there is a variegated group of humans loosely connected by an affinity for cryptocurrencies. This ranges from mere hobbyists to people working normal jobs and doing crypto as a side hustle to actual billionaires whose only sense of work is logging onto Binance and Bybit. This article will provide a look into my framework for portfolio management, sizing, and general participation within the crypto markets… but first, a little history of last cycle.

1. What goes up

From early 2020 to late 2021, we saw the market take BTC from $3,000 to $69,000 (23x), ETH from $80 to $4870 (61x), and also saw the rise of SOLUNAVAX from near nothing to multiples in the 3 or 4 digits. Now you’d think in a market like that, nearly everyone came out with a lot of winnings. They took profit, right? Not exactly. The collapse of the cryptocurrency industry in 2022 was just as spectacular as its rise over the previous years, as an already overbought and overleveraged market got cold water poured on it by Fed hikes, quantitative tightening, and token unlocks.

The collapse was so widespread that in the same way one could have thrown a dart at a board and longed whatever it landed on the previous two years, navigating your way out was akin to traversing an old minefield. You had to avoid:

1. sitting in majors which fell 80-85%

2. LUNA/UST, which vaporized mid 11 figures market cap to ~0 in about 4 days

   

3. Celsius, BlockFi, 3AC, Genesis, Gemini Earn, the GBTC/ETHE “arb” which left tons of dead bodies

4. FTX, which appears to have been SBF’s personal piggybank

   

5. blowing up or chopping yourself to death on perps

6. several other things that blew up that I couldn’t possibly list

2. Framework for Market Participation

With a little bit of history behind us, I want to talk about portfolio management, risk management, and sizing in a general sense but also in the context of the crypto market.

Every trade has a couple things in common that you should define.

1. Source(s) of edge- why is it a winning trade?

2. Conviction- how sure are you in [the size of] your edge

3. Target- this can be time-based, price-based, or event-based. Signals when you can exit

4. Invalidation- what things would allow you to conclude your idea was wrong and allows you to exit the trade

5. Range of outcomes/variance- how often will you win on the trade, how much will you win or lose in each case

6. Sizing- how much can you bet on the trade (in $, and in % of your portfolio)

7. Correlations to other bets and overall market- how does your bet relate to direction of the overall market and to your other bets

8. Tax consequences- how much tax do you pay (or save) when you win (lose). If none, good for you, because this is hard.

3. Don’t Blow Up

In a market that offers such a breadth of opportunities, the biggest key is to survive. For the context of a short, beginner-friendly article, I will assume you are not using leverage or if you are, it is very low. Introducing leverage is necessarily more complicated as it introduces issues around path dependence, capital efficiency (or lack thereof), and hedging costs (in some cases, positive carry!). So now you have 100% of a portfolio to allocate across a market with thousands upon thousands of ideas. You should always seek to work within your own skillset, portfolio size, and goals. It’s on you to be honest with yourself about what those are. Now, I’ll briefly expand on each of the 8 topics above.

Edge

As an ex-high-stakes-poker guy, this is instinct, like breathing. If you don’t have an edge, you shouldn’t be at the table. Wait for your spots, you are not here to try to get lucky. The best trades often have multiple sources of edge. Maybe you found an API leak or onchain traces that reveal an imminent Binance listing— edge. Trades with multiple sources of edge are preferred because they give you more ways to win. You might be wrong about one of them, or mistime it somewhat, either due to your own misinterpretation or unforeseen circumstances, but having multiple sources of edge will give you a better chance of winning regardless.

At every decision node (in both trading and poker), you have the opportunity to make the correct play. Sometimes you find yourself in a great spot on the turn in a big pot and are met with the worst possible river and a big bet, and the right decision is to fold. The same goes in trading. In the short run there is a lot of noise and you may lose money on ideas that were good the same way that you may lose when you were dealt two aces in Hold’Em. But edge over time dominates all, and keeps your equity curve (or bankroll) increasing.

Conviction

This is key as it relates a lot to sizing. Since your edge in a given trade is more intuitive and usually can only be estimated, this creates uncertainty bands. More on this ahead.

Target and Invalidation

Time-based, price-based, and event-based. These are all pretty straightforward but I’ll give a few examples. When Blackrock filed their BTC Spot ETF application this past summer, the price of BTC was ~$25,000. This was a great entry for the ETF trade given, among other things, Blackrock’s 99.8% approval rate. If you had a price based target, you may have chosen the significant multi-year level of $30,000-$32,000. If you had a time-based target, you may waited for the January 8-10 anticipated approval window and gotten to sell between $45,000 and $49,000. If you had an event-based target, you may have simply waited for the decision squawk and then used limit orders to get out.

Range of Outcomes, Variance and Sizing

Below is the formula for the Kelly Criterion, which defines the optimal bet size to maximize log(wealth), or the expected geometric growth rate. Note: I’ll leave it to the reader to define what their portfolio size is, whether you consider it to be your total wealth, total liquid wealth, or some amount that you have specifically set aside for crypto.

Without getting into the merits of using the Kelly criterion in practice (hint: don’t), there are two basic intuitions to glean: your bet size is proportional to your edge and inversely proportional to its variance. A very high edge, low variance bet leads to a large bet size. A low edge, high variance bet leads to a small (or 0) bet size, with the other two cases being somewhere in the middle. Expanding more now on conviction, this is a key third dimension because you often can only estimate your edge, and the probability distribution may be muddled by Black Swans or kurtosis, wherein the fat left tail is something like a DeFi protocol you’re using for a strategy getting exploited that results in total loss of funds. Bet size ought to scale with conviction, but it may also be constrained by market liquidity. A lot of the best, highest edge strategies are capacity-constrained and don’t allow huge bet sizes. Some examples of this may be a $5MM “lowcap gem” if you’re running a $50MM portfolio, or anything that relates to market microstructure.

Correlations

Without turning this into one of my old Statistics, Linear Algebra, or Econometrics textbooks, there is a lot of correlation within the crypto market. A rising tide (BTC, mostly) lifts all ships. Therefore, if you have a short thesis in a particular token, it may make sense to pair it with a corresponding long leg. Maybe you are looking at the upcoming Arbitrum cliff. You could choose to express this outright by shorting ARB, or you may express it as a pair by both shorting ARB and longing a comparable token like OP (or ETH, or some basket of correlated tokens). You can also look at correlations between totally different bets in your portfolio. Maybe, for example, you had a thesis around the GBTC unlock being bearish BTCUSD (see below), and you also had a thesis around TIA staking for the combination of staking yields, price appreciation as more people enter this trade, and future airdrops.

In this case, if your BTC short works out, there’s a good chance your spot TIA bags will also go down and vice versa, so these serve to balance each other out somewhat which reduces the volatility of your portfolio. It’s worth noting that the crypto ecosystem itself has a lot of dependencies and different layers of risk, whether that be protocols that compose with each other or the potential for a bug in the Solidity compiler.

Tax Consequences

This is a topic that is rarely discussed and very often misunderstood. At the end of the day, this is simply math — something that can be modeled and calculated rather than argued over. If you are American, you probably know short-term capital gains rates top out at 40.8% after the NIIT, while long-term capital gains rates are lower and top out at 23.8%. This leads to some weird distortions— I will show one below.

This is a theoretical multi-leg trade with the following assumptions:

-you enter and exit at the prices as shown (for simplicity, this factors in trading fees/slippage)

-the path of the price is as shown

-your STCG and LTCG rates are 40.8% and 23.8% respectively

In the top scenario, you manage to long ETH at $1,000, sell it at its local top of $2,000, buy it back at its local low of $1600 (20% pullback), and sell it again at the next top for $4,000. In this scenario, you trade several moves clairvoyantly (reality is, your trading will be worse!), but you also do not hold your positions long enough to reach the lower LTCG rate.

In the bottom scenario, everything is the same but you don’t trade the local move. Because of not trading the local move, you do hold your position long enough to reach LTCG. Perhaps counterintuitively, you end up with more money post-tax than in the previous scenario where you “successfully” captured the local move.

Now, this is best understood as a toy game with a specific set of assumptions, and is meant to illustrate that tax distortions may have a larger effect than you realize. There is nuance around spot vs futures and offsetting positions across different instruments. I assert that any set of assumptions can and should be modeled. There are lots of nuances and edge cases in these like timing (at what point in the tax year are you exiting) and yield considerations, but it helps to develop some basic intuitions around the financial mathematics of getting in, staying in, and getting out of your positions. This also reinforces the idea that when in doubt, you should zoom out. I know you guys love to trade the 1 minute chart, but most tax regimes have advantages for longer holding periods, and even if they don’t you still save money on trading fees/slippage by not going in and out.

4. Five (or so) Great Ideas

So with all of that said, how do you go about constructing a portfolio? Start by writing down all of your ideas in a spreadsheet. Then start filling in each of the numbered points above. What are your sources of edge and how confident are you in the size of them? How much can you bet on this? What timeframe will you stay in this trade, will you exit based on price, time, or some event? How will you know when you’re wrong and need to cut it? How does it fit into your overall portfolio and do you want to do the trade outright or with a hedge? And last, how do your taxes work when you win or lose on this trade?

Once you’ve done that, you should start to know which are your best ideas that have some combination of: large edge (maybe multiple sources), low(ish) variance, very high conviction, clear targets/invalidations, and sufficient market liquidity. If you have 5 or so great ideas that allow you to bet 10-30% of your portfolio each, that more or less allows you to allocate your full portfolio. The more good/great bets you have, the better. The more uncorrelated (or inversely correlated) they are, the better, as you’ll experience much more shallow drawdowns which can matter a lot in such a volatile market. You can then fill in various smaller bets around these bigger bets. Ideally these bets are just as good as the top ones, but maybe they only allow 1-5% of your portfolio. This is a constant process that never completes, as in the course of your ideation, you may find a trade that is better than your existing bets, and need to free up capital for it. Then you can simply wait for your targets or invalidations to come, and continue brainstorming new opportunities as trades play out and the capital allocated to them needs a new home.

5. Rope

One thing about trading/investing that is really great is you may see a million opportunities, but you get to choose when and how you risk money. You can allow the low opportunity, or low confidence ideas to simply pass by untouched, while reaching out and grabbing only the best ones. In that vein, you want to leave yourself some rope. In such a crazy market, a few times a year we see extreme dislocations and often all that is needed to capture them is your presence and free capital. The more times you do this, the more it will become like muscle memory— the opportunity presents, and you know exactly how to pounce on it. It’s good to have some free capital so you can fire big into such opportunities while they last. If you are fully allocated, it’s good to know which strategy or strategies you can free up capital from (lowest edge, not locked or trapped on an L2) in case one of these spots arises. Often times you need to act fast and don’t have time to agonize over how you’re going to free up capital or wait for a 1-7 day bridge.

A few examples I can remember were when a well-overcollateralized FEI was trading down at $0.70, when the UST peg shattered and LUNA entered its death spiral, and when FTX’s balance sheet leaked in early November 2022. These were all opportunities that allowed you to slam large bets which played out rather quickly for extremely high returns. A huge chunk of your PnL in trading comes from just these opportunities so you want to make sure of two things: you don’t miss them when they come around, and you bet the correct size when they do. You can’t do that without access to capital, so always leave yourself some rope!

*Disclaimer: Nothing here constitutes investment advice.

---