Last updated: 9 August 2026
How The Microcap Minute Global decides what appears in an issue, written out so a reader can check the work rather than take it on trust.
There are two steps and only one of them is mechanical. That distinction runs through everything below, so it is worth stating first: the tests applied to a company are fixed and repeatable, and the choice of what goes into those tests is a human reading of a chart.
One thing: how a sector, a fund or a company is moving against the S&P 500.
Not whether it is cheap. Not whether the business is good. Not what it will do next. A company can appear in an issue while losing money, and several have.
Everything else in an issue, the margins, the returns on capital, the growth, is reported so a reader can judge the business for themselves. It plays no part in selection.
An issue works downward: the eleven US sectors, then the industry funds inside them, then the companies.
Every week we read a Relative Rotation Graph of the eleven Select Sector SPDR funds against the S&P 500, on charts from StockCharts.com.
A rotation graph places two measurements on one picture.
Across is strength against the benchmark. Up is whether that strength is building or fading. Both are centred on 100, which means behaving exactly like the index. So the picture divides into four:
| strength building | strength fading | |
|---|---|---|
| stronger than the index | Leading | Weakening |
| weaker than the index | Improving | Lagging |
Things tend to travel clockwise through those four over time, though nothing obliges them to, and plenty stall or reverse halfway.
A tail is the path over recent weeks, and the arrowhead is the only thing telling you which end is now. Without it a tail is ambiguous in both directions, and reading one backwards turns a name rotating out of Leading into an apparent leader.
Four principles, each learned by getting it wrong first:
It is a screener, not a timing tool. It says where attention is going, not when to act, and nothing in an issue is an instruction to do anything.
On every ticker we look for a long tail. Trajectory matters more than position, so we favour names moving north east, in the Leading or Improving quadrant.
A name already leading but turning down is not a candidate. That exclusion matters as much as the rule itself.
Beneath the sectors sit thirty two industry funds, each holding one line of business: semiconductors, regional banks, biotechnology, airlines, gold miners and so on.
They are read the same way, on the same graph, against the same benchmark. They are judged on their own record, not on whether the sector containing them was chosen, because a line of business can gain ground while the sector around it loses it.
Up to nine appear in an issue. Every one is then put through the same ten tests the companies face.
No company in an issue is taken from these funds. They are a reading in their own right, not a route to a name.
Inside each chosen sector we read the graph again on the companies themselves, and note the ones turning up against the index.
Only companies that already pass the ten tests below are plotted, so the chart carries around twenty readable names rather than eighty that could not be used anyway.
The names that survive are then sorted into the businesses they are actually in, and the strongest of each is kept. Without that step a list can run five regional banks and three laboratory companies, which is one story told eight times rather than eight companies.
Every company and every fund must clear all ten. No score, no weighting, no exceptions.
keep = all([
price > ma_50, # above the 50 day average
ma_50 > ma_150, # the 50 above the 150
ma_150 > ma_200, # the 150 above the 200
rsi_weekly >= 50,
rsi_monthly >= 50,
rsi_daily <= 85, # not already overbought
price >= low_52w * 1.25, # 25% off the one year low
price >= high_52w * 0.85, # within 15% of the high
change_3m - spy_3m > 0, # ahead of the S&P 500
dollar_volume >= 10_000_000, # $10m a day
])
Given the same figures these return the same answer every time. The same threshold applies to a fund as to a company, including the liquidity bar, which has kept funds out of issues.
Each issue also reports the stage of a chart in Stan Weinstein’s sense: Stage 1 a base, Stage 2 an advance, Stage 3 a top, Stage 4 a decline. It is worked out from how far price sits above each moving average. A stage says where a chart is, not what to do about it.
| Step | |
|---|---|
| Reading the graphs, choosing sectors, funds and candidate companies | judgement, by eye |
| The ten tests | mechanical |
| Grouping by business and keeping the strongest of each | mechanical |
| Every figure reported | taken from the data, not chosen |
Another reader looking at the same graphs could reasonably reach a different list. That is a property of the method, not a defect in it, and it is why nothing here is presented as a rule that produces a right answer.
All of these are available in the data and none of them appears in an issue:
The next scheduled results date is printed, because a date is a fact rather than a forecast.
Each company card also shows its P/E against its own three, five and ten year averages. Three averages cannot give a range, because averaging destroys the spread a range would have to be recovered from: an average of twenty could come from a series that ran ten to thirty or one that never left nineteen to twenty one. What they do show is whether a company is dearer or cheaper than it has been, and in which direction that has been moving.
Read a very large average as a warning rather than a valuation. A single period of near zero earnings sends the ratio to hundreds and drags the average with it, so a high average usually marks a bad year rather than an expensive share. Where a company has no P/E today, because it is losing money, the history is left blank rather than shown against nothing.
StockCharts.com provides the Relative Rotation Graphs. They are read there on a paid subscription, captured by hand, and reproduced under their reprint permission policy.
StockAnalysis.com provides every figure we print. Two exports are downloaded by hand each week, one covering US stocks and one covering funds.
Nothing is read out of a chart into our text, tables or cards. The graphs decide what we look at, not what we print about it.
Stated plainly, because a method described without them is a sales pitch.
Relative Rotation Graph and RRG are registered trademarks of RRG Research, used with permission. The graphs are published by StockCharts.com and are read there on a paid subscription. Neither is affiliated with this publication and neither has reviewed or endorsed anything in it.
The full Disclaimer applies to every issue.