Dollar-cost averaging

You commit to an amount and a cadence, not to a forecast. Vanguard's February 2023 study found that deploying a lump sum beat averaging it in over 12 months roughly two thirds of the time, the honest case against the method. The case for it is that most people have income rather than a lump sum, and the ones with a lump sum often never press the button.

How it works

The mechanic behind dollar-cost averaging is arithmetic, and it works in your favor without requiring anyone to be right about the price.

Fix the money, not the quantity. A 100 dollar order buys 0.002 bitcoin at 50,000 dollars, 0.001 at 100,000, and 0.00125 at 80,000. Three purchases have spent 300 dollars and accumulated 0.00425 bitcoin, so the average cost per coin is 70,588 dollars. The simple average of the three prices you paid was 76,667. The gap exists because a fixed budget automatically buys more units when the price is low, and it appears in any series where the price moves at all.

What averaging does not do is reduce your exposure over time. Every purchase adds to a position that faces the same asset risk as a lump sum would, and after a year of weekly buys most of your money has been in the market for months. It smooths the entry price and it removes the decision, which is the part that actually defeats people. It is not a hedge, and it does nothing about the possibility that bitcoin is worth less in ten years than today.

The discipline part is real. Averaging only works if you keep going through the periods when continuing feels stupid, and those periods are long. Bitcoin spent most of 2022 below 20,000 dollars and did not trade above 100,000 until December 4, 2024. A schedule that survives that stretch does the job; one abandoned in month four does not.

Why this matters when you buy bitcoin

Three practical things decide whether an averaging plan works, and none of them is the price.

The first is fees, because they are charged per purchase and a small purchase makes them enormous in percentage terms. River charges no platform fee on recurring orders once the schedule has been running seven days, though a buy spread of roughly a quarter of a percent still applies. Swan is about 0.99 percent per purchase and Relai around 1 percent. Route the same schedule through a simple buy widget instead and you can pay a flat 2.99 dollar fee on a purchase in the 50 to 200 dollar band, close to 4 percent on a 75 dollar buy, plus the spread. On a weekly plan that difference compounds 52 times a year.

The second is custody. Averaging into an exchange account is averaging into counterparty risk, and the balance grows quietly for years if nothing forces you to look at it. Set a withdrawal threshold at the start, or pick a platform that pushes coins to an address you control automatically.

The third is tax record keeping, which people discover late. Every purchase is a separate lot with its own cost basis and its own holding period, because the IRS has treated virtual currency as property since Notice 2014-21. Fifty-two buys a year means 52 lots to match against any future sale, and US brokers now report disposals to the IRS on Form 1099-DA. Export the transaction history annually rather than reconstructing it under pressure.

100 dollars a week for a year

A year of weekly 100 dollar buys shows what an averaging plan actually costs.

Fifty-two purchases of 100 dollars puts 5,200 dollars to work across a full year of prices. At Swan's roughly 0.99 percent that costs about 51 dollars in platform fees. At a 2.99 dollar flat fee per purchase it costs 155 dollars, three times as much, before any spread. At River's zero-fee recurring rate the platform fee is nothing after the first week, and the buy spread becomes the cost you are actually paying.

Withdrawals are the other line. Moving each week's purchase to your own wallet separately multiplies whatever your exchange charges by 52, so the sensible pattern is to accumulate and sweep quarterly. Four withdrawals a year keeps the coins in your control on a timescale that matters, without paying for the privilege every seven days.

Dollar-cost averaging vs a recurring buy

Dollar-cost averaging is a decision about how to deploy money; a recurring buy is the button that carries it out. You can average manually, placing a limit order yourself every payday, and many people who care about maker rates do exactly that. You can also switch on a recurring buy for something that is not averaging at all, such as sweeping a variable bonus. Keeping the two separate is useful, because when the schedule underperforms the problem is almost never the strategy and almost always the fee tier the button quietly signed you up for.