dcahold

The Bitcoin return triangle: CHF 1,000 a month into BTC

Wed, 30 Sep 2026 00:00:00 GMT

Inspired by the return triangles of the Deutsches Aktieninstitut, here’s the same idea for Bitcoin: a theoretical version of my own savings plan, started in each of the past 10 years. Each cell is a plan that starts in January of the row’s year and is valued at the end of the column’s year.

The triangle is deliberately theoretical:

CHF 1,000 a month, no transaction costs. The full amount buys BTC.
Bought at each month’s closing price, on the last day of the month.
Valued at the closing price at the end of the exit year. Like the rest of the blog, this post updates with each new purchase, so prices run to September 30, 2026, the month-end of the latest transaction, and 2026 is a year-to-date figure. The Portfolio page has today’s values.
Launched in 2025. The row of plans started in the year the actual plan was launched is outlined in orange.

Each cell shows the PnL in CHF, value minus everything paid in, with the annualized money-weighted return (IRR) in brackets, the same reading as the Deutsches Aktieninstitut triangles. The cell color follows the return in brackets. One-year plans annualize a few months of average holding time, which inflates the diagonal, so the colors cap at ±100% a year. PnL grows with the amount paid in: a plan running from 2017 has paid in up to CHF 117,000, one starting in 2026 only CHF 9,000 so far.

IRR vs. PnL-based return. The IRR is the constant yearly rate at which every CHF 1,000 would have to grow, from the day it was paid in, to reach the plan’s value. It accounts for timing: later contributions had less time to grow. The PnL-based return, PnL divided by everything paid in, is a total return over the whole plan, not per year. Long plans therefore show a much larger total return than IRR, while short ones can show a larger IRR, because a few months are scaled up to a year. As of September 30, 2026, the 2017 plan is at 637.2% in total, 39.2% a year as IRR; the 2025 plan is at 1.5% in total, 1.8% a year as IRR.

How the actual plan differs

The theoretical version of the actual plan, from January 2025 to September 30, 2026, is at CHF 313 (1.5%), shown in the outlined 2025/2026 cell. The September 30, 2026 row of the full transaction log shows −CHF 68 (−0.3%) for the real purchases. The percentage here is the total return, the same measure as the log’s PnL (%), not the annualized return shown in the triangle. The actual plan differs from the theory in four ways:

Step	Difference	PnL (%)
Theoretical plan		CHF 313 (1.5%)
Amount paid other than CHF 1,000	+CHF 34	CHF 347 (1.7%)
Fees and spread	−CHF 389	−CHF 43 (−0.2%)
Settlement date and Kraken price per purchase	−CHF 39	−CHF 82 (−0.4%)
Valued at Kraken’s close	+CHF 13	−CHF 68 (−0.3%)

Amount paid other than CHF 1,000. January 2025: CHF 866, settled on February 3, 2025. Where the amount paid isn’t known, it assumes 2% costs on the BTC received.
Fees and spread. Every purchase pays fees and a spread, estimated at 1.8% of the CHF paid, or CHF 375 in total. The estimate is the CHF paid minus the value of the BTC received at that day’s closing price. The theory has no costs. In the table, the fees count at what the BTC they would have bought is worth on September 30, 2026, −CHF 389.
Settlement date and Kraken price. The actual purchases settle on or around the last business day of the month, and the log prices them at Kraken’s BTC/CHF close on that day. The theory uses the calendar month-end and Bitstamp’s BTC/USD close converted at the ECB’s USD/CHF rate.
Valued at Kraken’s close. The log values the BTC at Kraken’s September 30, 2026 close of CHF 69,794, the theory at the September 30, 2026 close of CHF 69,750.

Methodology and caveats

The triangle’s BTC/CHF price is Bitstamp’s daily BTC/USD close (UTC) multiplied by the European Central Bank’s USD/CHF reference rate. Kraken’s BTC/CHF prices, used for the transaction log, only go back about two years.
The log’s fees and spread are estimated from the CHF paid, the BTC received and Kraken’s daily close, not taken from the provider’s actual execution price and fee.
Taxes are ignored. Holdings are valued at the market price, with no sale and no selling fees.
Ten years is a short history for an asset this volatile, and the earliest entry years are dominated by a few large bull markets. Past returns are not a reliable indicator of future returns.

DCA transactions so far: entry price, BTC bought, and portfolio value

Mon, 31 Aug 2026 00:00:00 GMT

Since the savings plan went live, every incoming transaction to the wallet has been logged with the CHF paid for it, fees included, and matched to a historical BTC/CHF price. This is a short snapshot of where things stand and what changed with the latest contribution. It updates with every new purchase, using numbers as of the latest transaction. The full transaction log lives on the Portfolio page.

Where things stand

As of September 30, 2026, 22 transactions have gone in since February 2025. BTC/CHF rose from CHF 63,443 to CHF 69,794 (+10.0%) since the previous transaction, lifting PnL by +CHF 1,790. Portfolio value increased by +CHF 2,790 to CHF 20,798. Since inception, PnL is down −CHF 68 (−0.3%).

Metric	Value	Change since previous transaction
BTC/CHF	CHF 69,794	+CHF 6,351 (+10.0%)
Average entry price	CHF 70,023 per BTC	+CHF 33
Cumulative BTC bought	0.298 BTC	+0.01415 BTC
Cumulative CHF paid	CHF 20,866	+CHF 1,000
of which fees and spread (est.)	CHF 375	+CHF 12
Portfolio value (CHF)	CHF 20,798	+CHF 2,790
PnL	−CHF 68 (−0.3%)	+CHF 1,790 (+9.0%)

Portfolio value and PnL are marked at each transaction’s own date, so the “value” row reflects the BTC/CHF price on September 30, 2026, matching that row of the full transaction log. The Portfolio page’s key metrics use today’s price instead.

CHF paid includes fees: every purchase is CHF 1,000; the first one’s amount isn’t known and assumes 2% costs on the BTC received. Fees and spread are estimated as the CHF paid minus the value of the BTC received at that day’s closing price, so the average entry price is the break-even price after costs.

Cumulative CHF paid vs. portfolio value

When the blue line is above the grey line, the position is up overall; below, it’s down — this is about the running total, not any single purchase. See the Portfolio page for today’s live value and the full transaction-by-transaction log.

Setting up my Bitcoin savings plan

Mon, 20 Jan 2025 00:00:00 GMT

In order to implement my simple strategy of DCA, HOLD and Diversify (more on the reasoning behind it on the about page) I set up a Bitcoin savings plan. This post originally went up before the first purchase, when there was no track record to speak of yet — it’s been rewritten since, more than a year into actually running the plan, to state the rules plainly rather than mixing them with the numbers. The actual track record now lives in its own posts, linked below.

The plan

Automate it. The purchase happens on a fixed schedule, once a month, without me deciding in the moment whether it’s a good time to buy. Taking that decision out of my hands is the point — DCA only stays unbiased by market timing if it’s actually automatic, not just “usually monthly.” I run the recurring buy through Pocket Bitcoin, which handles the scheduled purchase itself. If you sign up there, feel free to use my referral code 37SNW2 — it gets you CHF 5 in Bitcoin once you’ve bought CHF 500 worth.

CHF 1,000 before fees, every month. A fixed amount, fixed currency, fixed cadence — no adjusting up when the price looks like it’s dipping or down when it’s ripping. “Before fees” is the number I’m committing to; whatever the exchange takes off the top is a cost of running the plan, not something that should quietly shrink the contribution.

Keep track of performance on a regular basis. Automating the buying doesn’t mean ignoring the result. Checking in periodically — how much has been spent, how much BTC that bought, what it’s worth now — is what turns “I set up a savings plan” into an actual record, and it’s the only way I’ll know if the plan needs revisiting.

Where the numbers live

This post stays about the rules, not the results. The actual purchase log, average entry price, and portfolio value are tracked separately on the Portfolio page, which shows the live values. Monthly commentary on what’s changed lives in the blog, starting with DCA transactions so far.

Why I’m writing this down

Mon, 16 Dec 2024 00:00:00 GMT

I’m about to commit to a Bitcoin savings plan — buy consistently, hold through the noise, don’t put it all in one basket — and I’m writing it down in public instead of tracking it in a private spreadsheet. A spreadsheet is easy to quietly edit when the numbers get uncomfortable; a dated blog post is not. That’s the whole reason this exists: accountability, not advice.

The full strategy is spelled out on the about page. This is day zero — no purchases yet, no results, just the intent to follow through and let the record speak for itself.