I initially drafted this post in April 2024, but didn’t publish it… here it is with a post-script for my activity since then…

Sizing and Portfolio Management: Retail’s Advantage

As a trader or investor, one of the most important factors that affects returns is position sizing. Position sizing is often one of the first things capital markets participants learn about and the last thing they master. Everyone who has participated in the market is familiar with the feeling “when I’m right it’s too small and when I’m wrong it’s too large”. This can be true both psychologically (because losses hurt more than wins) and in reality (if we size up late to chase a move).

Inexperienced traders/investors often look to learn from institutional players who are playing a fundamentally different game. Institutional position sizing is classically “up to 5%” of the book for a single position. This is partly because these larger institutional players are often trying to beat a broad benchmark/index vs. drive absolute returns. Additionally, when running an institutional book, a 2% position could be $100mm+, and putting that position on (or unwinding it) may require many days so as not to move the market meaningfully. Smaller participants do not have these constraints and should treat this as a feature, not a bug.

For capital markets participants that are playing the total return game, a 2-5% position is irrelevant and leads to the feeling of “I was right, but it doesn’t matter”. On a $250k retail book, a 2% position is $5k — if that doubles, “wow I crushed that trade”, but it was only worth 2% of your equity. But how common is a 100% trade? Answer: not as common as a 20-30% move.

Thus, if we can size up to 10% of our book and hit a 20% move (which happen much more frequently), we can match the outcome of a double on 2% of equity. The downside is obviously that our exposure is 5x higher when we use 10% positions vs. 2% positions — so, when sizing up investors need to use tighter stops. And that is where our ability to unwind positions quickly without moving the market comes into play.

Retail traders should use their relatively smaller size to their advantage. Concentrating on highly liquid markets, small players can easily put on a 10%+ position without moving the market at all.

How I Approach Position Sizing

1) Scan weekly charts to identify trends and set ups that look appealing. We want to see charts that have tightened up over 3-6+ weeks and are above key moving averages on the daily chart, I use the 21 & 89 EMAs.

I generally trade breakouts, which does not work in all market conditions. Knowing when to de-gross and just watch the market is probably more important than finding good setups. An easy way to identify market conditions is the observe the weekly and daily trend of the SPY and it’s component sectors (XLK, XLC, XLE, etc).

2) After identifying stocks/indices that are set up well and confirming that I am trading with the trend (EG: XOM set up will and XLE in a trend or also set up well), I will open a “pilot position” which is generally 3% of my equity. I want the trades to start working immediately, and if so, will look to add on additional short-term bases. See Bill O’Neil annotated charts to understand the “pyramiding” process of adding “progressive exposure”, there are many available online and I have linked to one here. “Winners average winners, losers average losers”.

I always have a hard stop loss of 8% on any position because this is the upper limit before the law of compounding begins to work against you (i.e. 8% down, 8% up get you to flat but 10% down requires 11% up to get you to flat). But when using pilot positions, I generally look for the market to confirm my position by moving in the expected direction within 1-2 days. The best trades work right away.

THAT BEING SAID

I have been allocated 100% to IBIT (or 95/5 IBIT/MSTR, or 90/10 IBIT/ETHA as of today) since about May 2024 (BTC price ~$60k-$65k). About the time I was writing the original article it dawned on me that my “best” idea was IBIT and until that changed there was no reason for me to be trading other vehicles. I decided not to publish the article b/c it was contradictory to my PA…

IBIT will not be my best idea forever, but a large chunk of my portfolio rolls into long term capital gains treatment at the end of Jan 2025… based on the thesis described in my other posts, I think the crypto market tops some time between April and Sept/October 2025. So my goal is to get as much of my brokerage account into LTCG tax treatment and then sell 30-50% when I think we are frothy.

Signs of froth:

1. Bloomberg/CNBC running stories on crypto with a positive connotation: Not happening

2. Buddies that know I follow the market texting me about the best crypto to be in: Happened initially when BTC > $100k, but not happening this time

3. Barber/Uber driver/tradesmen talking about crypto: Not happening

4. Parents or their friends talking about (or asking if they should be) buying Bitcoin: Somewhat happening - they actually said I should have told them to buy, but now it’s “too high”

I plan to use my “retail advantage” when it is time to exit… if hedge funds or other institutional capital allocators want to exit IBIT or ETHA, it is going to take them days or weeks. As retail participants, we can exit in one day over 3-5 trades and not move the market at all.

As Promised on Twitter

Here are a few more Solana AI coins that I am closely watching. Currently, my only two AI coin positions are $ARC and $NEUR- I may buy more or sell the whole thing at any time for any reason.

$ARC has a very tight chart and I have rolled my other crypto AI positions into this coin. My understanding is that ARC is playing a VIRTUALS-like game whereby the ARC coin needs to be purchased to fund the initial liquidity pool for any AI coins launched through their platform. With these new crypto “metas”, I try not to focus too much on fundamentals, instead following the money flow.

$NEUR is a lowcap AI coin focused on creating the scaffolding for ai agents to interact within the Solana ecosystem. As of this writing the coin is at $70m market cap. The project currently has integrations with Jupiter, pump fun, Magic Eden, and Dex Screener and works with Claude 3.5 Sonnet and GPT-4. The project is in closed beta and available to access for 1 SOL.

$GRIFFAIN this platform allows anyone to launch their own “agent”. GRIFFAIN is interesting in that the agents have their own wallet, can crawl Twitter/X, and have some access to crypto data. The agent can perform tasks at will, or tasks that the user sets in advance, i.e. “when XYZ coin hits a market cap of $100m, buy 1 SOL”.

Cheers to 2025 - Back Soon.