The question « which is the most expensive ? » calls for a maximum, and a maximum is an announcement: that of a seller, on a given day.As of September 6, 2026, the highest listing in the 1986 series was a certified first issue at $6,000, in a segment whose median was $250 across nineteen listings. The second highest was a fourth certified issue at $4,900, while the median for its segment was $88 — fifty-six times less. These two figures do not describe a market hierarchy; they describe two particularly optimistic salespeople.
What a maximum measures, and what it does not measure
In an ad dataset, maximum is the most fragile value of all. The median is based on half of the observations, and moving an ad hardly moves it. The maximum comes down to a single line: remove it, it changes completely.
This fragility would be acceptable if the maximum designated something solid. This is not the case here, for a reason relating to the very nature of the source: an ad expresses what a seller wishes to obtain. Nothing obliges it to be realistic, nothing penalizes it if it is not, and an overpriced ad can remain online indefinitely without anyone being upset.
The maximum of an announcement statement therefore measures the highest ambition of the day, and it is information - but not the information you think you are reading. It provides information on the dispersion of sellers' expectations, never on the top of the market.
The highest announcements recorded on September 6, 2026
In descending order, the observed peaks are distributed as follows. The first issue peaked at $6,000 certified, out of nineteen ads whose median was $250. The fourth reached $4,900 certified, out of nine listings with a median of $88. The twelfth capped at $600 certified on three listings, the second at $550 on sixteen, the first at $315 non-certified on forty-three, the third at $300 non-certified on one hundred and nine, the fourth also at $300 non-certified.
A connection is necessary between these peaks and the corresponding medians, because it is this relationship which reveals the nature of each figure. On the fourth certified fascicle, the maximum is worth fifty-six times the median. On the first certified, twenty-four times. On the second certified, eight times. On the third uncertified, twenty-three times.
A ratio of fifty-six to nine observations is not the sign of a two-story market: it is the sign that one announcement in nine is very far from the other eight. Statistically, this is called an outlier, and it is customary to point it out rather than headline it.
$6,000 — the top of the statement
A first certified booklet. Its segment median was $250 and its floor was $89. The announcement is therefore twenty-four times the median, in a segment that is nevertheless well-stocked with nineteen observations.
$4,900 — the most revealing
A fourth certified booklet, in a segment of nine ads whose median was $88 and the floor was $30. Fifty-six times the median: no other figure in the statement better illustrates what a maximum is worth as an indicator.
$315 — the peak of uncertified
A first booklet without certification, in a segment of forty-three ads at a median of $60. Only five times the median: the raw segment, more abundant, produces much less extravagant peaks.
$600 on the twelfth — to be discarded
Three certified advertisements in total on this booklet. The maximum of a set of three values is not a market maximum, it is simply the highest of three listings.
The defensible hierarchy of this series is not read in the peaks but in the medians accompanied by their numbers. It fits in one line: the first issue at $60 on forty-three ads, then all the others between $10 and $30, the tenth bringing up the rear of the non-firsts at $30 on twenty ads.
This hierarchy is boring and robust. That of maximums is spectacular and dissolves as soon as a seller withdraws his ad.
What would actually make a booklet expensive?
Discarding outliers does not remove the fundamental question. What, in this series, could make a number rise sustainably?
Three conditions are usually combined, and the survey allows us to test two of them. The first is the scarcity of supply. It exists here, but only at the end of the series: three ads on the twelfth, twenty on the tenth and eleventh, compared to 146 on the second. The second is a floor that does not give way, a sign that no holder is selling off. The tenth issue is the only one to present it, with a minimum of $19 when most of the others go below six dollars.
The third condition is sustained demand, and this is one that an ad survey cannot observe. We see what is offered, never what is purchased. Two of the three conditions are therefore verifiable and one is not, which prohibits any conclusion - but indicates where to look if we want to follow the question over time.
The tenth fascicle brings together the two observable conditions. It's the only one in the series in this case, and it's a more interesting fact than a $6,000 ad, even if it's less well told.
Titrer sur le maximum
The highest figure in a reading is also the least representative. It attracts attention and tells nothing other than the person who stared at it.
Confondre affichage et transaction
A $4,900 listing sets no price. It can stay online for years; its presence in a statement says nothing about its conclusion.
Estimate your copy on a peak
A holder who aligns his price with the observed maximum finds himself alone at the top of a range whose median is fifty times lower.
Ignorer l'effectif du segment
The maximum of three ads and the maximum of one hundred and nine do not compare. The first is one of three values, the second is really the peak of a distribution.
Assign value to certification
The peaks are in the certified segment, but because the rating there varies enormously. The case does not add value: it reveals that of the copy.
Treat a summit like a record
A maximum number of ads is not a sales record. The two words designate facts of a different nature, and only the second implies that a transaction has taken place.
The top of the raw segment, more informative than that of the certified
A detail goes unnoticed and deserves attention, because it reverses intuition. The highs of the non-certified segment are much lower than those of the certified segment: $315 on the first issue, $300 on the third and fourth, $125 on the sixth, $94 on three others. Compared to their medians, these maximums represent five to twenty-three times, compared to twenty-four to fifty-six times on the certified side.
The explanation lies in what each segment measures. In the raw, all the copies are described by their sellers with vague vocabulary, and a potential buyer can compare one ad to a hundred others. A price that is too high is immediately apparent, and the seller who deviates too much is left alone. Competition naturally tightens the top of the range.
In the certified, each copy carries a distinct note, and two announcements of the same issue can legitimately differ by a factor of ten. This legitimacy creates a space where a very high price does not seem absurd: it could correspond to an exceptional rating. The certified segment therefore offers shelter from extreme values that the raw segment does not provide.
Practical consequence: the maximum of the raw segment is a more honest indicator than that of the certified, because it has been confronted with the comparison. In this series, the highest price asked for an ordinary copy of the first issue was $315, and this figure describes the market better than the $6,000 posted for an encapsulated copy.
The rule generalizes beyond this series. In any segment where objects are directly comparable to each other, extreme values are contained by competition from neighboring advertisements. In any segment where each object has an attribute that singles it out, extreme values become unverifiable and multiply. Knowing which of the two we are in determines the credit we give to the observed peak.
How to read a ranking of the most expensive
Rankings of this type circulate widely, and it is possible to use them provided you ask them three questions before believing them.
The first: are these prices asked or sales concluded? Both exist, they cannot be compared, and the gap between them is systematically to the disadvantage of the first. A classification that does not specify this probably mixes the two.
The second: how many observations is the figure based on? A peak from three announcements and a peak from one hundred do not have the same status. The absence of numbers is the most reliable sign that a ranking has not been constructed, but compiled.
The third: on what date? A statement without a date is unusable, not because it is false, but because it cannot be verified or compared to anything. This is why each amount in this article bears its own.
A ranking that answers all three questions deserves consideration. It will be less spectacular than the others, and that is precisely what makes it usable. The details by booklet appear in ouranalysis of the first series.
The highest listing in the September 6, 2026 statement was a first certified fascicle at $6,000. This amount is a price asked by a seller, in a segment whose median was $250 across nineteen listings. It does not constitute a sales record or a market level.
This is the question that must be asked, and the statement does not answer it. What we know: the certified segment of this booklet had nine ads, its median was $88 and its floor was $30. This announcement is therefore fifty-six times the median of its own segment.
Almost certainly not. Aligning your estimate with the observed maximum amounts to placing yourself at the top of a range. The usable benchmark is the median of the segment corresponding to the actual state of the specimen, accompanied by its number.
The tenth is the only one to meet the two observable conditions: a restricted supply, twenty advertisements, and a floor which does not give way, at $19. The third condition, sustained demand, cannot be measured by a survey of announcements, which prevents us from drawing a forecast.
By checking three points: are these prices asked or sales concluded, on how many observations is each figure based, and on what date was the statement made. A ranking that does not exhibit these three elements has been compiled rather than constructed.
The median and its number, not the peak
My Comics Collection records each estimate with its full range and the number of advertisements on which it is based, so that the benchmark retained is the one that describes your copy.