Japan Stock Files Free · unsigned · primary sources
Fourteen Documents, One Recurring Number · Fragment 14

The One-Point Band

SubjectHikari Tsushin 9435 HIKARI TSUSHIN, INC.

Fourteen large-shareholding documents, fourteen different issuers, eleven days. Twelve of them print a before-and-after ratio, and ten of the eleven increases land between +1.00 and +1.08 points. The per-filer columns show what the summary line is made of.

17 primary sources · figures as-of 24 Jul 2026 · ¥/USD 163.8 method · Corrections: 0

A variation report numbered 11 was filed on 7 August 2026 on Risk Monster Co., Ltd. (3768, Tokyo Stock Exchange). The date on which the reporting obligation arose is printed as 31 July 2026. In the summary table, two figures sit in adjacent columns: 16.10% now, 15.10% in the previous report — a difference of exactly 1.00 point.

The filer is Hikari Tsushin, and it lodged thirteen more large-shareholding documents in the eleven days ending 7 August, each on a different issuer. In order of filing: Human Creation Holdings (7361), variation report No. 8, obligation 21 July, filed 28 July. Quest (2332), an initial large-shareholding report, 22 July, filed 29 July. Japan M&A Solution (9236), No. 15, 22 July, 29 July. PCA (9629), No. 3, 22 July, 29 July. NTT Data Intramart (3850), No. 3, 23 July, 30 July. Just Planning (4287), an initial report, 23 July, 30 July. Medical System Network (4350), No. 5, 23 July, 30 July. System Location (2480), No. 8, 27 July, 3 August. J.S.B. (3480), No. 16, 27 July, 3 August. Freebit (3843), No. 29, 29 July, 5 August. Commerce One Holdings (4496), No. 6, 29 July, 5 August. Paraca (4809), No. 2, 30 July, 6 August. Nihon O.A. Kenkyusho (5241), No. 7, 31 July, 7 August. And Risk Monster, 31 July, 7 August. Every issuer here is listed on the Tokyo Stock Exchange, except Nihon O.A. Kenkyusho, whose report prints the Nagoya Stock Exchange; System Location's prints both.

Fourteen obligation dates, fourteen filing dates, and one interval: seven calendar days, in every one of the fourteen. The statute sets the outer edge of that distance. A large-shareholding report is due within five days of the day the holder became a large holder, and a variation report within five days of the change — 五日, a count from which Sundays and the holidays a Cabinet Order names are excluded. Every one of these fourteen obligation dates is a weekday, and every one of the fourteen filing dates is the same weekday of the following week: seven calendar days on, which in this window is the fifth counted day, the last the count allows.

Twelve of the fourteen print a before-and-after pair in the summary table. Risk Monster, 15.10% to 16.10%. Nihon O.A. Kenkyusho, 11.34% to 12.37%. Paraca, 6.02% to 7.05%. Freebit, 16.77% to 17.30%. Commerce One Holdings, 9.49% to 10.49%. System Location, 11.73% to 12.76%. NTT Data Intramart, 7.36% to 8.44%. Medical System Network, 9.18% to 10.20%. Japan M&A Solution, 18.86% to 19.86%. PCA, 7.07% to 8.08%. Human Creation Holdings, 11.77% to 12.77%. J.S.B., 18.64% to 0.00%. The other two documents are initial reports, on Just Planning and on Quest, and the prior-ratio column on an initial report has nothing to carry; it is printed blank.

Of those twelve pairs, eleven rise and one falls. The eleven rises: +1.00, +1.03, +1.03, +0.53, +1.00, +1.03, +1.08, +1.02, +1.00, +1.01, +1.00. Ten of them land between +1.00 and +1.08 points. The eleventh is Freebit, at +0.53.

Each variation report prints a reason field. Eight of them read the same two clauses: 株券等保有割合の1%以上増加、単体株券等保有割合の1%以上増加 — an increase of 1% or more in the holding ratio, and an increase of 1% or more in the single-filer holding ratio. That is the text on Nihon O.A. Kenkyusho, Paraca, Commerce One, System Location, Medical System Network, Japan M&A Solution, PCA and Human Creation. Two thresholds, printed on one line, and the number in both of them is one.

Three read otherwise. Risk Monster's reason field names only the single-filer clause, while its summary pair moves 15.10 to 16.10, a difference printed as exactly 1.00. Freebit's also names only the single-filer clause, and its summary moves +0.53. Intramart's says only that the holding ratio rose by 1% or more, without naming a single filer, and its summary moves +1.08. For Risk Monster and Freebit, the line that explains the filing and the line that reports the group are not counting the same thing.

The per-filer columns are where that can be checked, because a document of this type prints a before-and-after ratio for each filer and not only for the group. Risk Monster's four printed pairs run 7.30% to 7.30%, 7.33% to 7.33%, 0.18% to 1.18%, and 0.29% to 0.00%. One further line has no prior figure at all. Add the prior side of the four: 7.30 + 7.33 + 0.18 + 0.29 = 15.10, the printed prior. Add the current side: 7.30 + 7.33 + 1.18 + 0.00 = 15.81, and the line with the blank prior supplies 0.29 to reach the printed 16.10.

FIG. 6 — MIXED-UNIT TRANSMISSIONSCALE NOT COMPARABLEUse: parent-sub control line① Parent input shaft: 3,000 RPM (revolutions per minute)② Coupling: type unlisted · ratio not on file③ Sub. output shaft: 4 RPM (resolutions/meeting)Note: RPM differs by shaftRev. 1994 · torque figures withheld① Parent shaft (rpm=rev/min) ② Coupling ③ Subsidiary shaft (rpm=res/mtg)
FIG. 6 — Transmission diagram, parent-subsidiary control line: both shafts rated in "RPM," neither meaning the same thing.

The breakdown table names the current holders with their share counts: Hikari Tsushin, 557,200 at 7.30%; UH Partners 2, 559,800 at 7.33%; UH Partners 3, 90,300 at 1.18%; UH6, 22,100 at 0.29%. The counts sum to 1,229,400, which is the printed total, and the ratios sum to 16.10%, which is the printed ratio. Four names, which is also the count of filers and joint holders printed at the front; the line that falls to 0.00% is not among them, and UH6 arrives at 0.29% without a prior column, so in the total the two offset. What remains of the group's +1.00 is UH Partners 3 moving from 0.18% to 1.18% — the one leg the reason field names.

The construction repeats. Nihon O.A. Kenkyusho prints legs of 7.56% and 0.75% unchanged and one leg from 3.03% to 4.06%; the prior legs sum to 11.34% and the current legs to 12.37%, both matching the summary line, so the group's +1.03 is that one leg's +1.03, against a joint holding of 198,900 shares. Paraca prints 0.16% unchanged and one leg from 5.86% to 6.89%; the sums are 6.02% and 7.05%, again matching, on 736,200 shares. Again the group difference is a single filer's difference.

Commerce One: 7.52% unchanged, one leg from 1.97% to 2.97%, summary 9.49% to 10.49%, legs summing exactly on both sides, 754,400 shares. PCA: 6.85% unchanged, one leg from 0.22% to 1.23%, summary 7.07% to 8.08%, exact on both sides, 1,777,800 shares. Human Creation: 6.63% and 0.35% unchanged, one leg from 4.79% to 5.79%, sums of 11.77% and 12.77% against the printed pair, 456,400 shares. Counting these three with Nihon O.A., Paraca, System Location, Medical System Network and Japan M&A Solution, eight of the eleven rising documents move exactly one filer's line and reprint every other line to the hundredth. Nothing is being aggregated in these tables so much as copied upward.

A ratio printed to two decimals is a rounded number, and a sum of rounded numbers is not the same object as a rounded sum. In five of the twelve, the printed legs and the printed summary line sit 0.01 apart, on one side of the pair or the other: Japan M&A Solution, System Location, Medical System Network, Freebit and NTT Data Intramart. Japan M&A Solution prints 7.48%, 7.67% and 1.12% unchanged and one leg from 2.59% to 3.60%. The current legs sum to 19.87% while the summary line reads 19.86%; the prior legs sum to 18.86%, which matches. So the leg moved +1.01 and the summary moved +1.00. The total is 301,800 shares, expressed once, as 19.86%.

System Location carries the residue on the other side. Legs of 7.46%, 0.15% and 0.19% unchanged, one from 3.92% to 4.96%. The current legs sum to 12.76%, matching the summary. The prior legs sum to 11.72% against a printed prior of 11.73%. The moving leg is therefore +1.04 while the summary difference is +1.03, on a joint total of 455,700 shares at 12.76%. Which of the two a reader wants depends on which line of the form the reader is reading.

Medical System Network carries it on both sides at once: 7.32% and 0.37% unchanged, one leg from 1.50% to 2.52%, the legs summing to 9.19% and 10.21% against printed figures of 9.18% and 10.20%. Because both sides sit 0.01 away in the same direction, the difference survives intact at +1.02, which is exactly the moving leg. The joint total is 3,125,700 shares at 10.20%. The residue is visible in the levels and invisible in the change.

Freebit is the fourth of the five, and it is also the one rise outside the band. It prints three legs: 6.16% unchanged, 8.08% to 7.60%, and 2.53% to 3.55%. The moving legs are −0.48 and +1.02, netting +0.54 against a printed difference of +0.53; the prior legs sum to 16.77% exactly and the current legs to 17.31% against a printed 17.30%, on 4,050,900 shares. The +0.53 in the summary is a one-point leg net of a half-point leg, plus a hundredth of rounding.

Intramart sits at the top of the range of rises, at +1.08, and it is two filers. The first prints 7.36% to 7.51%, a move of +0.15. The second has no prior column, and its current ratio is printed twice: 0.92%, on 45,800 shares against 4,955,000 shares outstanding as of 23 July. The summary prior, 7.36%, is the first filer's ratio alone. The two current legs sum to 8.43% against a printed summary of 8.44%, which makes this the fifth document carrying a hundredth of residue. No single leg here moves a point: the first moves +0.15, and the second is a whole holding of 0.92% standing where there was no prior column at all. The highest of the eleven rises and the lowest are two of the three documents in which more than one line differs from the prior report.

A filing of this kind does not measure a position. It measures a crossing.

The one decrease is filed under a different limb of the same provision. Its 【根拠条文】 reads 法第27条の25第1項及び第2項, where the other eleven variation reports cite the first paragraph alone; the second paragraph applies where a holder has transferred a large volume in a short period, and requires the transferee and the consideration to be reported as well. Variation report No. 16 on J.S.B., filed 3 August against an obligation date of 27 July, prints three legs at 6.77%, 7.19% and 4.68% before and 0.00% after, summing to the printed 18.64% and 0.00%, on a holding of 0 shares. Its reason field lists four items where most of the others list two: a decrease of 1% or more in the holding ratio, the same for the single-filer ratio, a decrease in joint holders, and a change in a material contract such as a pledge.

The two initial reports print a dash in the reason field, there being no earlier report to differ from. Just Planning, filed 30 July against an obligation date of 23 July, two filers, 612,900 shares against 12,229,888 outstanding, 5.01%. Quest, filed 29 July against 22 July, one filer, 275,100 shares against 5,487,768, also 5.01%, and a blank prior column. The obligation begins where the ratio exceeds five per cent — 百分の五を超える, in the statute's words — and both of these print the first hundredth above it. An initial report starts where a variation report's arithmetic cannot: the entry in the before column is not zero, it is nothing at all, and the difference that the twelve other documents are built around is undefined here.

A point is a different quantity in each of these documents, and each of them prints the quantity it is. 発行済株式等総数 — total shares outstanding, with an as-of date — appears in every filer's section, and one per cent of it is a point. Japan M&A Solution, 1,519,300 shares outstanding, so 15,193 to the point. Nihon O.A. Kenkyusho, 1,608,000, so 16,080. System Location, 3,570,000, so 35,700. Human Creation, 3,573,100, so 35,731. NTT Data Intramart, 4,955,000, so 49,550. Commerce One, 7,193,400, so 71,934. Risk Monster, 7,637,154, so 76,371.54, the only one of the twelve listed here whose total is not a multiple of a hundred — the two initial reports print 12,229,888 and 5,487,768, which are not either. Paraca, 10,441,400, so 104,414. J.S.B., 21,982,000, so 219,820. PCA, 22,000,000, so 220,000. Freebit, 23,414,000, so 234,140. Medical System Network, 30,642,600, so 306,426. The leg that moves about a point is worth 15,193 shares in one of these documents and 306,426 in another, and both of them print as about a point.

Those quotients are built from the current side only, and that is the limit of what these columns support. The lines used here carry a ratio for each filer before and after, a current share count for each filer and for the joint holding, and a per-holder breakdown of the current joint ratio: on Human Creation, 237,000 at 6.63%, 206,900 at 5.79% and 12,500 at 0.35% against a joint 456,400 at 12.77%. Eleven of the twelve variation reports print per-holder rows in that table; J.S.B.'s prints the same three headings and a single total line, 0 shares at 0.00%. A prior share count per filer is not among the entries: the prior-side box on each filer's line is 「直前の報告書に記載された株券等保有割合」, a ratio and nothing else. The denominator is printed — 発行済株式等総数, with its as-of date — and so is a sixty-day record of acquisitions and disposals, dated and quantified: Intramart's second filer has ten rows between 9 July and 23 July, from 100 shares to 11,000, summing to 45,800, which is the whole of its holding. So a leg can be read against a printed denominator and a printed sixty-day record. What cannot be reconstructed from these columns is the same split on the prior side.

All fourteen documents print a 保有目的 field for every filer; the form provides one. In twelve of them every field reads 純投資 — pure investment. Four of Risk Monster's five carry it; the fifth is blank, on UH5's line, which reads 0.00%. J.S.B.'s three are blank, on a holding of 0 shares. In both, the blank sits on a line with nothing left to hold. What the documents establish is what the documents print.

So: fourteen documents, one interval of seven days, and the number one in three places. It is in the reason field, on eight of them twice over, as the threshold that makes the report due. It is in ten of the eleven rising summary lines as a difference between +1.00 and +1.08. And in eight of those eleven it is one filer's line moving from something to something-plus-one while every neighbouring line is reprinted unchanged. The eleven rises are narrowly spread because the threshold that selected them is the same number they report. A filing of this kind does not measure a position; it measures a crossing. Three of the eleven rises have more than one line differing from the prior report: Risk Monster, Freebit and Intramart. The two furthest from the middle of that spread, at +0.53 and +1.08, are two of the three; in Risk Monster the other two lines offset each other in the total.

A note on conversion, since this publication converts yen and this piece has none to convert. The lines quoted above carry share counts, holding ratios, dates and filer counts, and no yen amount appears among them. Where a yen figure does appear in this publication, it is converted at ¥163.8 = US$1, the rate retrieved from the source below on 24 July 2026, applied to every period regardless of when the yen figure was struck, with the significant figures of the original yen amount preserved. Ratios, share counts, filer counts and document counts are reported as printed and are not converted; the shares-per-point figures above are one per cent of a printed total, and are exact.

What would change our mind

Everything above rests on two printed columns and one interval, and all three are republished on a schedule. Each of these fourteen issuers will draw a further document from the same filer whenever a holding ratio crosses the threshold again, and each will print an obligation date, a filing date and a new before-and-after pair. If the next batch prints intervals other than seven calendar days, the constancy described here belongs to this eleven-day window rather than to the form. If the next increments scatter well away from a point while the reason fields still cite the one-percent clauses, then the threshold is not doing the selecting and the narrow spread was coincidence. And the shares-per-point quotients assume the denominator behind the unchanged legs held still: a later report whose joint total and joint ratio imply a materially different denominator would undo that arithmetic, and it would say so in the same two columns.

Sources

  1. Risk Monster (3768) variation report No. 11, filed 7 August 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YTTJFinancial Services Agency, EDINET
  2. Nihon O.A. Kenkyusho (5241) variation report No. 7, filed 7 August 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YTTOFinancial Services Agency, EDINET
  3. Paraca (4809) variation report No. 2, filed 6 August 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YTTHFinancial Services Agency, EDINET
  4. Freebit (3843) variation report No. 29, filed 5 August 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YTADFinancial Services Agency, EDINET
  5. Commerce One Holdings (4496) variation report No. 6, filed 5 August 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YTAFFinancial Services Agency, EDINET
  6. System Location (2480) variation report No. 8, filed 3 August 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YSP6Financial Services Agency, EDINET
  7. J.S.B. (3480) variation report No. 16, filed 3 August 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YSP7Financial Services Agency, EDINET
  8. NTT Data Intramart (3850) variation report No. 3, filed 30 July 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YSB5Financial Services Agency, EDINET
  9. Just Planning (4287) large-shareholding report, filed 30 July 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YSB7Financial Services Agency, EDINET
  10. Medical System Network (4350) variation report No. 5, filed 30 July 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YSB9Financial Services Agency, EDINET
  11. Quest (2332) large-shareholding report, filed 29 July 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YSB1Financial Services Agency, EDINET
  12. Japan M&A Solution (9236) variation report No. 15, filed 29 July 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YSB2Financial Services Agency, EDINET
  13. PCA (9629) variation report No. 3, filed 29 July 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YSB4Financial Services Agency, EDINET
  14. Human Creation Holdings (7361) variation report No. 8, filed 28 July 2026 — https://disclosure2.edinet-fsa.go.jp/WZEK0040.aspx?S100YS3EFinancial Services Agency, EDINET
  15. Japanese yen exchange rate — the live quote page the ¥163.8 rate was retrieved from on 24 July 2026. It shows the current rate, not that date's; the reference-rate source below holds the dated value — https://tradingeconomics.com/japan/currencyTradingEconomics
  16. Financial Instruments and Exchange Act, Articles 27-23(1) and 27-25 — the five per cent threshold, the five-day filing period, the 1% change that makes a variation report due, and the additional disclosure required of a short-period large transfer — https://laws.e-gov.go.jp/law/323AC0000000025e-Gov, Digital Agency
  17. USD/JPY for 24 July 2026 — 163.82 on a dated reference series blended across central-bank sources. This URL keeps returning that date's value, so the rate behind every dollar figure here can be checked after the fact; the conversions use ¥163.8 — https://api.frankfurter.app/2026-07-24?from=USD&to=JPYFrankfurter (a dated series blended across central-bank sources)
Every figure above is traceable to the source it came from, and every source is linked. If one of them is wrong, write to [email protected] — we correct within 72 hours, at the top of the piece, and leave the history public.

This is general information about public disclosure, published freely to an unspecified readership. It is not investment advice, and contains no rating, target price, or recommendation.