A stock index rises 4% while an industry group gains 2%. Both charts point upward, but the group is losing ground relative to the index. John Murphy’s work encourages precisely this wider view: the direction of one chart is only part of the information available.
StockCharts’ tribute to Murphy, who lived from 1942 to 2026, describes his work as Chief Technical Analyst and his emphasis on intermarket analysis and the primary trend. His StockCharts biography also identifies relative strength and sector rotation as important themes.
The following examples are hypothetical comparisons, not Murphy’s trades or a reproduction of his rules. They show how an analyst can use broader context without turning a relationship between markets into an automatic forecast.
John Murphy and the choice of comparison
Take the opening example with both series starting at 100. The index finishes at 104 and the sector at 102. The sector-to-index ratio falls from 1.00 to approximately 0.9808, a decline of about 1.92%. The sector has a positive absolute return and a negative relative return.
I find it useful to read those two results side by side, because 2% growth can look strong on an isolated chart while the 0.9808 ratio shows the lag. The ratio shows exactly what the isolated chart leaves out. Neither result has to be discarded to make the other true.
A different benchmark could change the comparison. A technology group might outperform a global index while lagging a domestic growth index. Each benchmark asks a different question, so changing benchmarks after seeing the answer can make almost any chart look like a leader.
For related background, see the site’s sector rotation and relative strength guide. Relative strength in this setting means comparative performance. It doesn’t mean the Relative Strength Index oscillator, which describes gains and losses within one price series.
Four charts can describe one crowded exposure
Imagine a portfolio containing an energy producer, an oil-services company, a commodity fund and a currency position. Four labels may suggest diversification. Yet a single hypothetical decline in energy prices could affect several of them at once.
A simple scenario starts with four equal positions, each worth 10,000. If three lose 8% and the fourth is unchanged, the combined loss is 2,400, or 6% of the 40,000 total. Counting holdings wouldn’t reveal that concentration of outcomes.
The scenario assumes no fixed relationship between oil and any currency. It asks the analyst to list plausible common sensitivities and then examine whether the data support them. The same group of instruments might behave differently during a supply disruption, a demand shock or a change in financing conditions.
A useful intermarket note can distinguish the observed move from the proposed explanation. “These three assets declined this week” is an observation. “They declined because of the same cause” requires additional evidence. An attractive economic story can be wrong even when the chart comparison is accurate.
The observation window can reverse the message
Suppose a sector gains 12% over six months but falls 6% during the latest month. An index gains 8% over six months and loses 2% during that latest month. The sector leads over the longer window while lagging over the shorter one.
A trader studying a multi-month trend might retain the longer comparison while noting the recent deterioration. A short-term researcher might investigate the last month’s reversal. The record becomes confusing when the chosen window changes each time the preferred conclusion comes under pressure.
For example, a worksheet could always show one-month, three-month and six-month relative returns. The three windows are independent illustrative choices. Their purpose is to expose the sensitivity of the interpretation, rather than to prescribe a Murphy formula.
A simple moving average offers another way to describe trend over a fixed lookback. An average smooths data; it doesn’t remove the consequences of choosing that lookback. Two analysts can disagree because they’re measuring different intervals, even if both calculations are correct.
Correlation doesn’t establish a leading signal
Two markets can move together without either providing advance notice of the other. If both respond on Tuesday to the same announcement, their relationship may be visible after Tuesday’s close. A strategy entering before the announcement couldn’t have used those completed returns.
Consider a hypothetical test that uses today’s bond-market close to explain today’s equity-market return. It might find a useful association for description. Calling it a tradable prediction would require an entry time at which the bond information was already available and the relevant equity return still lay ahead.
I would mark that information timestamp explicitly. A value recorded at 16:00 can’t justify a decision made at 10:00 on the same day. A plausible cross-market story doesn’t repair that six-hour gap.
Relationships can also change. A correlation estimated from the previous sixty observations describes that sample. It doesn’t promise the same sign or strength in the next sixty. The analyst needs a response to disagreement that was specified before the disagreement arrived, whether that response is further review or a defined change in the research model.
Currency can change the apparent winner
A hypothetical overseas asset gains 10% in its local currency while that currency loses 8% against the observer’s home currency. Ignoring costs and distributions, the combined home-currency return is 1.10 multiplied by 0.92, minus one, which equals 1.2%.
Adding 10% and minus 8% would give 2%, an inaccurate result because the changes compound. The distinction can matter when comparing domestic and overseas charts or when interpreting a global sector ranking.
A chart displayed in local currency may still be appropriate for studying the asset’s local market behaviour. An account measured in another currency faces a different return. The units belong in the comparison alongside the date range.
For another Masters profile, see Jim Rogers. Cross-border research raises a practical question: what is the exposure actually being measured, and in which currency?
A useful filter needs its rejected cases
Suppose a hypothetical breakout study records twenty signals. Ten occur when a chosen sector ratio is rising and ten when it is falling. The analyst can compare both groups with the same entry, exit and cost assumptions. Merely displaying the strongest examples from the rising-ratio group would leave the filter unevaluated.
There is a trade-off to measure. A filter might avoid some failures while also excluding the largest winner. It might reduce the number of trades without improving their average result. A neat story about supportive context can’t decide which of those outcomes occurred.
Small groups are especially sensitive to a single observation. If nine results total minus 3R and one gains 8R, the group finishes at plus 5R. Removing the largest winner changes the conclusion sharply. R here means the study’s fixed unit of initial planned risk.
The dependence on one winner makes the distribution worth examining beside the average. Wider samples, changing market conditions and realistic execution assumptions remain relevant before a descriptive relationship becomes part of a trading method.
Keep disagreement visible on the worksheet
A stock can close above 80.00 resistance while its sector ratio weakens and its benchmark remains below an earlier high. That combination supplies conflicting observations. An analyst can preserve those disagreements while examining each observation on its own terms.
A research record might preserve all three facts, then classify similar cases consistently. If a planned response changes, the new rule receives a date. Otherwise it’s easy to explain away each disappointing result with whichever surrounding market happened to move adversely.
Murphy’s broad perspective is valuable at this point. More charts give the analyst more context, but they also create more opportunities to choose a convenient explanation. Keeping the comparisons stable makes the extra information easier to question.
Books for studying John Murphy
These four books are by John J. Murphy. Technical Analysis of the Financial Markets updates his earlier Futures Markets text. Intermarket Analysis and The Visual Investor are labelled by edition below, while Charting Made Easy offers a shorter entry point. The 102-to-104 comparison above is an independent calculation, not a quoted example from any of these books.
As an Amazon Associate, I earn from qualifying purchases.
Technical Analysis of the Financial Markets. Hardcover.
(US) Technical Analysis of the Financial Markets
(UK) Technical Analysis of the Financial Markets
(DE) Technical Analysis of the Financial Markets
Intermarket Analysis. Paperback, second edition.
(US) Intermarket Analysis (Paperback, second edition)
(UK) Intermarket Analysis (Paperback, second edition)
(DE) Intermarket Analysis (Paperback, second edition)
The Visual Investor. Hardcover, second edition.
(US) The Visual Investor (Hardcover, second edition)
(UK) The Visual Investor (Hardcover, second edition)
(DE) The Visual Investor (Hardcover, second edition)
Charting Made Easy. Paperback.
A relative chart earns a precise caption
“The sector is strong” leaves too much open. “The sector gained 2% while its benchmark gained 4% over the same month” gives the reader a comparison they can check. Murphy’s wider view becomes practical when every additional chart brings that level of clarity.
Learn the pattern. Ride the trend. Keep the gains.
Educational content only. Not investment advice. Trading involves risk. You are responsible for your decisions.
Get the free Market Wisdom e-book
Join Trends and Breakouts — historical winners, breakout studies, and risk lessons. No spam, unsubscribe anytime.