Should I use a spreadsheet or an app to track my target allocation?

Three tabs are enough for most people. The layout below is one reasonable design, not a standard.

Tab 1: Targets. One row per fund or asset class.

(Funds are examples only, not recommendations.)

Tab 2: Holdings. One row per fund.

Tab 3: Contribution. This is where the work happens.

For column E, there are two cases. If the sum of all needs is at least the contribution, split in proportion to need: =D/SUM(D)*$B$1. If the sum of needs is smaller than the contribution, every fund first gets its need and the extra is split by target percentage: =D + ($B$1-SUM(D))*Target%. Many first sheets forget the second case.

A fourth tab, a plain log of date, ticker, shares and price, gives you history.

Check your sheet with a known answer. Feed it this example: a $50,000 portfolio with VTI at $33,000, VXUS at $13,500 and BND at $3,500, targets 60/30/10, and a $1,000 contribution. It should return VTI $0, VXUS $529.41 and BND $470.59 (the steps are in How much to invest each month). If it does not, you have found a bug before it cost you anything. The same test works on any app.

You can type prices, or pull them. Google Sheets has GOOGLEFINANCE, which Google documents as "Real-time price quote, delayed by up to 20 minutes," and says the data is "not for trading purposes or advice." Excel's Stocks data type uses LSEG Data & Analytics (formerly Refinitiv), and Microsoft says the content is provided for information purposes only and is not intended for trading. Prices may be delayed, so check Microsoft's notice for the exchange you care about.

For a monthly contribution a 15- or 20-minute delay rarely matters. A stale or missing price does.

Formulas, mostly. The research on spreadsheet errors is old and was done on business spreadsheets, not personal portfolio trackers. Treat the numbers as a warning about human error, not a forecast for your sheet.

• Researcher Raymond Panko's review of intensive inspection studies of operational spreadsheets found errors in 94% of the 85 spreadsheets studied.
• His earlier review (2000) points the same way: the most recent field audits at the time found errors in at least 86% of the spreadsheets audited, and the older audits used methods unlikely to catch most errors.
• Based on human-error research, he expects cell error rates in the range of 1% to 5%; lab studies of 967 people building spreadsheets averaged 3.9%.
• Errors are hard to catch: in experiments, people inspecting spreadsheets found an average of 60% of the errors.
• Panko's conclusion: errors are "rare on a per-cell basis," but in large spreadsheets at least one wrong bottom-line value is very likely.

Illustrative math. Suppose each of 30 formulas had a 1% chance of being wrong, the low end of that range, and the errors were independent. The chance that at least one is wrong is 1 − 0.99^30 ≈ 26%. A small allocation sheet has simple, copied formulas, so the real risk is likely lower, but it is not zero.

Common failure points in a rebalancing sheet are structural and easy to name (these are my own judgment):

• A row inserted outside the range of a SUM, so a fund drops out of the total.
• A price typed by hand and never updated.
• Targets that sum to 110% and nobody notices.
• Contribution formulas that ignore the case where needs total less than the contribution.
• Whole-share rounding left out, so the buy list asks for 7.56 shares of a fund your broker only sells whole (see Fractional shares and rebalancing).
• No history, so you cannot see what you bought last month.

• Total control. You can see and change every formula. If you want a rule no tool has, you can write it.
• Cost. If you already use a spreadsheet program, the added cost can be zero.
• Ownership. The file lives where you put it.
• Learning. Building the sheet forces you to understand the contribution math.

These are judgment calls. Consider an app if:

1. You hold more than a handful of funds, or use several accounts or currencies.
2. You need whole-share rounding handled, not just dollar targets.
3. You have skipped updates before. A tool that fails when you forget is worse than none.
4. You want history and charts without building them.

Allocta is one example. It is built around buying only: it calculates how to split a contribution so your holdings move toward their targets and it never tells you to sell. It is not a broker and does not place trades. Its free plan covers up to five holdings, and new accounts get a seven-day trial of the paid plan. Whatever you choose, ask what happens to your data, whether you can export it, and what the tool does when it has no price for a fund.

Less often than most people think. Vanguard's site says an annual rebalance is optimal for many investors and that the key is a schedule that is easy to follow. Fidelity lists calendar, threshold (for example 5 percentage points off target) and hybrid approaches, and warns against tinkering with a portfolio excessively. A tool you open twice a year can be a spreadsheet. A tool you open every month with each contribution leans toward an app.