Eight years ago I bought my first Bitcoin to win an argument with a friend. Today I run the crypto desk at Milkos, and the question I get most often from readers is the same one I asked myself in 2017: how much of my portfolio should this take?

The honest answer is "less than you think, sized so you can leave it alone." Here is how we get there.

Step 1 — Define what you are buying

For portfolio-construction purposes we treat only two assets as genuine candidates: Bitcoin and Ethereum. Everything else is either a private bet you are running on your own, or part of our crypto research notes — not part of an allocation.

  • Bitcoin — monetary alternative. We size it next to the gold sleeve.
  • Ethereum — platform exposure with cash-flow logic via staking. We size it next to growth equities.

Step 2 — Apply the survivability test

Before assigning any weight, we apply a single rule we call the survivability test: the chosen position must be small enough that a 60% drawdown does not force a behavioural mistake or break the household plan. For most readers that translates to no more than 7–10% combined in BTC and ETH.

It is a deceptively simple rule. It eliminates the people who are about to buy too much at the top — and gives the disciplined investor a clear ceiling to refill toward at the bottom.

Our default 2025 sizing

  • Conservative reader (retirement within 5 years): 0–2% Bitcoin only, no staking.
  • Moderate reader (20-year horizon): 3–5% Bitcoin + 2% Ethereum.
  • Growth-oriented reader (younger than 35): up to 6% Bitcoin + 4% Ethereum.
"The right crypto allocation is the one that lets you forget about it for a year and then rebalance." — Mihai, internal Q3 review note

Step 3 — Decide whether to stake

Staking ETH currently pays around 3.2% per year, gross. After validator commission, MEV variance, and the inevitable downtime, the net is closer to 2.8%. After Romanian income tax on the rewards, the after-tax yield for most readers is 2.1–2.4%.

That is meaningful but not transformative. We recommend staking only if you can hold ETH for at least three years, because the staked position is less liquid and the tax accounting is more involved. Otherwise the simpler unstaked position is fine.

Step 4 — Choose custody honestly

The biggest unforced error we see is custody mismatch: meaningful sums sitting on exchanges that were never designed to hold them long term. Our default split:

  • Up to €5,000 — keep on a regulated EU exchange for ease of trading.
  • €5,000–€25,000 — combine exchange access with a hardware wallet for the bulk.
  • Above €25,000 — multi-signature setup with one institutional custodian and a personal hardware backup.

The arithmetic is unforgiving: a single self-custody mistake at €25,000 costs more than every paid subscription you will ever buy.

The costs nobody wants to talk about

Even with disciplined allocation, crypto carries costs that quietly compound:

  • Spread and slippage on entry and exit, typically 0.05–0.30% per round trip.
  • Tax-event tracking — every transfer between wallets is reportable in Romania even if no profit is realised.
  • Volatility drag — rebalancing a 60% drawdown back to target is psychologically expensive even when mathematically right.

We model these explicitly in our portfolio template so readers see the net expected return, not the gross.

What this looked like through October 2025

The moderate template (4% BTC + 2% ETH) contributed +1.4 percentage points to total portfolio return year-to-date, with a maximum drawdown contribution of −3.1 points during a brief August correction. That is consistent with the design: meaningful upside, bounded downside, and no required attention beyond the annual rebalance.

One question to ask yourself

If your crypto sleeve dropped 60% next month, what would you do? If the answer is "panic," the position is too large. If the answer is "rebalance and forget," you have the right size. The whole framework collapses to that single question — and once you answer it honestly, the math falls into place.