Fundamental Analysis vs Technical Analysis

Michael Whitman · Trading since 2017, full-time since 2020 · Published 2026-09-08 · Updated 2026-10-19 · Reviewed 2026-10-04

The oldest argument in the subject, and mostly a false one. Fundamental analysis asks what a business is worth. Technical analysis asks what price is doing. Those are different questions, so the useful answer is rarely which is better. It is which question your horizon requires, and what you do on the days the two answers point in opposite directions.

What each one is

Fundamental analysis estimates worth from the business. Revenue, margins, cash flow, debt, and a judgment about the future turned into a number. Fundamental analysis covers the approach.

Technical analysis reads what price has done. Levels, trends, ranges and volume, on the argument that price already reflects what participants collectively know. Technical analysis covers it.

Neither claims to predict. The honest version of both is a framework for deciding under uncertainty, and both are wrong regularly.

Where they differ

Horizon. A valuation resolves over years. A level resolves over bars. Applying one on the other’s timescale is the most common way either gets misused.

Whether it produces an invalidation. A chart level gives a specific price at which the idea is wrong, which is what sets position size. A valuation does not: “cheaper” is not a stopping rule, and a cheap company can get cheaper for years.

What being right costs. Fundamental work can be right about the business and still lose money for a year, because the share price answers to more than this year’s earnings. The data section counts how often that happened to companies in the Dow today.

What the inputs are. One reads filings that arrive quarterly; the other reads a price that updates continuously. That difference in refresh rate is most of why the horizons differ.

What worth and price readings share

Both are backward-looking. Filings describe a period that ended; charts describe bars that closed. Neither has access to the future, and both are often presented as though they do.

Both require a written rule to be reviewable. A valuation with no stated assumption and a chart pattern with no stated definition fail the same way: neither can be shown to have been wrong.

And both are worthless without position sizing. Whichever framework produced the idea, the amount at risk comes from the same arithmetic: risk figure divided by distance to invalidation.

Which one to use: Fundamental Analysis or Technical Analysis?

Use fundamental work when your horizon is years and you can hold through being early. Owning a business for a decade needs a reason a chart cannot supply, and the drawdowns along the way are the price of that horizon.

Use technical work when your horizon is days or weeks. Over that window nothing in the accounts has changed, so the only information available is what price is doing, and you need an invalidation, which only the chart provides.

Use both when you hold for months, with fundamental work choosing what to own and technical work choosing when to buy it. That combination is common and it needs one rule.

That rule is: when they disagree, the shorter horizon wins on timing and the longer one wins on selection. Deciding it in advance is what stops the disagreement being settled by whichever answer you preferred.

Earnings went up, the stock went down

This is the case the argument is really about. A fundamental analyst who called the business correctly, earnings up on the year, and still watched the shares fall, was right about one question and lost money on the other.

It is not rare. Across current Dow members, diluted earnings per share rose in 314 company-years, and in 54 of them (17.20%) the share price ended the fiscal year lower. The median change in those 54 years was a fall of 8.73%, and in 10 of them the shares lost 20% or more.

The deepest was Disney’s fiscal year to 1 October 2022. Diluted earnings per share rose from $1.09 to $1.72 and the shares fell 46.41%. Goldman Sachs is the other kind of example: its 2018 diluted earnings per share were $25.27 against $9.01 the year before, nearly triple, and its shares fell 34.43% in 2018. A rising number on the income statement told you nothing about where the price would sit twelve months later.

This is where an invalidation earns its keep. A holder working only from the accounts had no price at which the idea was declared wrong, so a 46% fall was something to sit through or to rationalize. A technical stop would have ended the position at a stated loss, and whether that was the better outcome depended entirely on the horizon the holder had chosen.

The original data

The sample. The companies held by the SPDR Dow Jones Industrial Average ETF on 1 October 2026, minus Visa, which files no standard diluted earnings-per-share item in its SEC data. Today’s members only, so the set leans toward businesses that did well: companies dropped from the index along the way are not in it, and a test on them could look worse.

Earnings and prices. Diluted earnings per share from each annual 10-K via SEC EDGAR’s XBRL company facts, comparing each year with the prior year printed in the same filing, so stock splits cannot distort the change. Prices are Yahoo Finance closes, split-adjusted, dividends not added back, taken on the last session on or before each fiscal year end. Fiscal years end between September 2009 and July 2026.

What was counted. A company-year counts as “earnings up” when diluted earnings per share were higher than the year before as reported in that 10-K, and “price down” when the close at the fiscal year end was below the close a year earlier. No earnings figure in the sample was exactly unchanged. Of the 314 earnings-up company-years, 54 (17.20%) ended lower, with a median price change of -8.73%. The ten falls of 20% or more: Disney to October 2022 -46.41%, Salesforce to January 2026 -37.87%, Nvidia to January 2012 -37.25%, Goldman Sachs 2018 -34.43%, Nvidia to January 2019 -34.18%, Sherwin-Williams 2022 -32.61%, 3M 2022 -32.49%, IBM 2018 -25.91%, Microsoft to June 2026 -25.01% and JPMorgan 2011 -21.62%. Every company-year, with its accession number and filing date, is in the Dow earnings and price file.

When it fails

The failure is switching frameworks to avoid an exit, and it is available to anyone who knows both. A technical trade goes against you, the level breaks, and the stop should fire, but the company is cheap on the numbers, so the position becomes a long-term hold. Nothing was reassessed. A short-horizon trade with a defined invalidation was converted into a multi-year position with none, at the exact moment the original rule said to leave, and the count above shows that improving earnings are no protection over the next twelve months.

The second failure is applying a valuation on a weekly horizon. Nothing in the accounts changed.

A third is expecting a chart to tell you what something is worth. It has no such input.

A fourth is holding a fundamental view with no invalidation. Cheaper is not a stopping rule.

A fifth is no rule for disagreement. The preferred answer then wins.

And a sixth is treating either as predictive. Both describe what has already happened.

Related

Fundamental analysis covers the business side and the statements it reads. Technical analysis covers the price side and where its levels come from. And combining the two covers using both without letting either rescue the other.

This page is educational, not financial advice. Test every idea on your own charts before risking money.