Mike Webster: Testing Rules Against the Chart Record

A growth-stock screen returns twelve names. Three later double, and those three charts become the examples saved for study. Mike Webster’s interest in market history raises a more demanding exercise: preserve what the screen showed before those outcomes were known, including the nine less memorable names.

In the mid-year 2024 MarketSurge Stock Guide interview, pages 16 to 19, Webster discusses work with William O’Neil’s research beginning in 1997 and becoming a portfolio manager in 2000. The interview covers CAN SLIM, historical chart study, MarketSurge tools, personal style and reassessing rules.

Those themes support a profile about research and judgement. The examples below are independent hypotheticals, rather than a proprietary Webster box system or a claim that he always places stops beneath a particular range.

Mike Webster and the original twelve-name screen

Suppose the twelve names qualified on the same Friday. A useful study retains that dated list along with the settings used to produce it. Reconstructing the list six months later can introduce changes to fundamentals, rankings or the set of companies still covered by the data provider.

I would keep all twelve rows because the three eventual winners don’t answer the original selection question. If those winners alone remain in the notebook, the screen appears to have produced a much more certain result than it actually did.

The interview names tools including Pattern Recognition, Growth 250, Composite and relative strength measures. Their presence in the discussion doesn’t make any one output a completed trade decision. A candidate list still needs interpretation and a process for managing outcomes.

The William O’Neil profile provides related context for the growth-stock tradition. An association with that tradition doesn’t justify assigning every familiar chart rule to Webster. The evidence needs to support the specific attribution.

Judgement can have a recorded reason

Imagine two hypothetical candidates, each near 50.00 resistance. One has spent six weeks in a range from 46.00 to 50.00. The other reached 50.00 only yesterday after a rapid move from 38.00. A researcher might regard them differently even if a broad screen includes both.

The useful next step is to state the reason for the distinction. Range duration, depth and distance from a prior move are observable features. A label such as “better quality” is less useful unless the record explains what produced it.

Discretion can work with a visible process. A note made before the outcome can describe why one candidate received attention and the other didn’t. Later review can then ask whether that distinction helped across comparable cases.

The market structure guide offers language for describing those differences. The value of a six-week duration and the likely outcome for the second candidate remain separate hypotheses requiring evidence.

A chart replay has an information boundary

Suppose a company reports earnings on Thursday after the close, and a stock breaks above resistance on Tuesday. A replay evaluating Tuesday’s decision cannot use Thursday’s release as if it were already available.

Publication dates matter as much as the figures themselves. A database may show revised earnings beside an old chart without preserving when the revised number became known. A clean-looking retrospective screen can therefore include information unavailable to the original observer.

I mark Tuesday and Thursday separately in this example because the two-day gap changes the research. An attractive explanation for why the stock later advanced may still be unusable as evidence supporting the earlier decision.

The same issue applies to weekly bars. Friday’s completed high, low and close can’t be used to describe a Tuesday setup without looking ahead. A replay can instead use the partial week as it stood, or wait until the completed bar genuinely existed.

Relative strength needs the right meaning

In a hypothetical month, a stock rises 8% while its benchmark rises 12%. It has advanced in absolute terms and lagged in relative terms. Starting both series at 100 gives ending values of 108 and 112; the ratio falls to approximately 0.9643.

A comparative ranking can express similar questions across a universe of stocks. It is different from the Relative Strength Index oscillator, which uses the pattern of gains and losses within one price series. Each tool answers its own measurement question.

The relative strength and sector guide provides the relevant comparison here. A high rank also needs a date and an observation window. Rankings can change, and a later leading stock may have looked quite ordinary at an earlier decision point.

A leadership measure describes performance under a chosen comparison. Participant identity and future persistence require additional evidence. The researcher still needs to decide what additional evidence is relevant to the study.

A rule change deserves its own comparison

Suppose an existing hypothetical rule requires a close above 50.00. After missing a strong move, the researcher proposes allowing an intraday crossing instead. A session reaching 50.30 and closing at 49.70 qualifies under the new rule alone.

The new rule will create additional entries, some favourable and some adverse. Evaluating only the missed winner gives it an unfair advantage. A comparison can apply both definitions to the same historical period with the same cost and exit assumptions.

The original result remains useful as a baseline. If the new rule performs better in that sample, a later period can examine whether the improvement persists. Repeatedly adjusting the rule to the same examples can fit their accidents unusually well.

Webster’s interview supports reassessing rules. Each proposed change still needs an evaluation of its effect. The practical discipline is giving a proposed revision a date, a reason and a record that includes its new failures.

Position management remains a separate research task

Consider a hypothetical entry at 50.20 and an intended adverse exit at 48.20. The initial price distance is 2.00. With 150 shares, planned price loss is 300 before costs. A later fill at 47.20 would make that loss 450.

The numbers make a simple distinction: identifying a promising candidate doesn’t complete the trade design. Execution, position size and exit conditions still affect the outcome. A list of successful stocks can’t supply those missing decisions.

The ATR-based swing stop discussion is an independent modern comparison for examining planned distance. It’s an independent comparison for this example. Actual fills during an abrupt change can differ from levels based on recent movement.

A study can record the stock’s eventual advance and the simulated trade result separately. If the exit occurs before a later rally, the trade doesn’t inherit that rally merely because the original candidate selection was interesting.

Personal style needs an operational description

Suppose one hypothetical reader can review charts twice daily while another can monitor each fifteen-minute close. They may interpret the same price sequence differently because the second observer sees intermediate information the first misses.

A personal style becomes inspectable when it specifies the decisions, timing and information it requires. “I prefer to be flexible” leaves those requirements unclear. “I reassess after each completed daily bar” describes a schedule that can be compared with actual behaviour.

A defined schedule makes the exercise of judgement easier to review. A record can show whether a decision came from new information, a documented exception or an emotional reaction to the previous trade.

The twelve-name screen then becomes the beginning of a coherent research process. Each later choice can be traced to what was known at the time, including the candidates left untouched and the rules that were revised.

A dated decision record can preserve disagreements too. Suppose two readers review the same twelve names and select four each, with only two names in common. Their agreement is incomplete despite an identical starting screen. Saving the reasons for inclusion and exclusion allows a later comparison to examine judgement directly. Without those notes, the eventual winners can make either reader’s preferences seem more consistent than they were. The original screen and the later selections are separate stages, each with its own opportunity for hindsight.

Books for studying Mike Webster

How to Make Money in Stocks is William J. O’Neil’s book, identified in Webster’s interview as an influence; the links select its fourth-edition paperback. Trade Like an O’Neil Disciple is by Gil Morales and Chris Kacher. Monster Stocks is by John Boik. Boik’s book studies historical stock winners. These separate accounts can sit beside Webster’s interview while retaining each author’s own perspective.

As an Amazon Associate, I earn from qualifying purchases.

How to Make Money in Stocks. Paperback, fourth edition.

(US) How to Make Money in Stocks (Paperback, fourth edition)
(UK) How to Make Money in Stocks (Paperback, fourth edition)
(DE) How to Make Money in Stocks (Paperback, fourth edition)

Trade Like an O’Neil Disciple. Hardcover.

(US) Trade Like an O’Neil Disciple
(UK) Trade Like an O’Neil Disciple
(DE) Trade Like an O’Neil Disciple

Monster Stocks. Paperback, 2023 edition.

(US) Monster Stocks (Paperback, 2023 edition)
(DE) Monster Stocks (Paperback, 2023 edition)

Keep the other nine charts in the folder

The three eventual winners are worth studying. The other nine names preserve the uncertainty that existed when the screen first ran. A research habit shaped by Webster’s interview becomes more useful when the saved record includes both the opportunity and the limits of the original decision.

Learn the pattern. Ride the trend. Keep the gains.

Educational content only. Not investment advice. Trading involves risk. You are responsible for your decisions.