Skip to content
#small language model Open access

Run One Rule on Five Clocks and Only the Direction and the Cost Agree — The Direction Stays Below a Cost of 0.7123 to 0.7304 Pips on All Five and the Cost Differs by Only 1.0254 Times, Yet the Count Differs by 57.4639 Times and the Duration of a Bar by 62.5735 Times [F055]

Sep 2026 · Zenodo (CERN European Organization for Nuclear Research)

Abstract

足種は時計である。時計で刻む足、出来事の数で刻む足、変位の絶対量で刻む足がある。では同じ一つの規則をその五つで回したら、何が揃って何が揃わないのか。規則の文は五つとも一字も変わらず、変わるのは「一本」の意味だけである。だから答えが変わったとすれば、それは規則ではなく一本の意味が変えたことになる。本稿の分離子は、何で刻んだか(時計・出来事の数・変位の絶対量)である。 四銘柄(USDJPY・EURUSD・GBPUSD・XAUUSD)の2019年から2024年までのティックから五つの足を作った。T 一分足、H 一時間足、K ティック足(N ティックごと)、V 出来高足(出来高の合計が N になるごと)、R レンジ足(経路の長さの合計が S になるごと)である。N は 5,120・4,534・4,780・9,046 ティック、S は 511.6875・340.5750・488.2750・1749.0050 pips で、一時間足と本数がそろうように取ってある。規則は「直前の足の向き(終値 − 始値の符号)で建て、24 本先で決済」で、D は平均(向き × これから 24 本先の値動き)を pips で測り、t は日次クラスタで作った。コストはその足の中のスプレッドの中央、散らばりは上位一割の点 ÷ 下位一割の点である。基準線として、時刻とスプレッドを実測のまま置いた合成乱数(3 系列)を使った。 揃ったのは二つだけだった。方向の情報 D は -0.0540 から +0.2789 pips のあいだにあり、五つともコスト 0.7123 から 0.7304 pips を下回る。そのコストは最大 ÷ 最小で 1.0254 倍しか違わない。つまり時計を変えても払う額は変わらず、出てくる答えも変わらない。一方で揃わないものは大きく違う。件数は 57.4639 倍、一本の時間は 62.5735 倍、一本の値動きは 11.3276 倍ちがう。表は「1.0254 倍のコスト」と「数十倍の群」の二つに割れる。なお D の行そのものは符号をまたぐので比が定義できず「揃わない」に入れてある。揃っているのは「五つともコストを下回る」という点であって、D の数値ではない。そこは分けて書く必要がある。 一定にしたものが揃う、という対角の形も出た。一時間足の時間の散らばりは 1.00 で上位一割と下位一割が同じ値であり、ティック足の値動きの散らばりは 14.92 でいちばん小さい。逆に時計で刻む足は値動きが散らばり、一分足 32.25・一時間足 20.63 である。ティック足と出来高足は件数も D も t もコストも同じ値で、切れ目の一致率は 1.0000 だった。これは為替では出来高がティックの本数だからで、検算であって発見ではない。一分足だけ日次 t が 2 を超えた(-2.9864)が、額は -0.0540 pips で、一本の値動きの中央 0.7250 pips がまるごとコスト 0.7129 pips に消える足である。件数が 2,084,015 あるので小さな額でも t は立つ。t が立つことと額が足りることは別である。 そして予言④は外れた。変位で刻むレンジ足の 15.01 より、出来事の数で刻むティック足の 14.92 のほうが小さい。理由は、レンジ足が一定にしているのが経路の長さであって一本の正味の値動きではないからである。足の中で上下に往復すれば経路は伸びるのに正味の値動きは残らない。つまり「値動き」という一語が、往復で増える量(経路の長さ)と往復で消える量(正味の変位)という二つの別の量を指しており、私はその二つを区別せずに予言を書いた。同じ言い方の誤りは本系列の別稿でも踏んでいる。ただし 14.92 と 15.01 の差はほとんど無く、「外れた」の中身は「レンジ足が負けた」ではなく「差がつかなかった」である。合成乱数でも D は -0.1598 から +0.0222 pips、日次 t の絶対値は 0.1044 までで、同じ形になった。結論として、時計を選ぶことは答えを選ぶことではない。時計を選ぶと件数と一本の意味が変わるだけで、出てくる答えは変わらない。「この足種なら見える」と言うときに要るのは、足種の名前ではなく、その足で測った額とコストを並べた一行である。 この論文の位置づけ:新しい数学定理も新しい法則も主張しない。時間変形と従属確率過程は古典的な題材である。特定の教材やその著者、販売元は主題ではなく、一切名指していない。規則を一つに固定しており、出口を論じていない(固定の先までの値動きで測っている)。成績も損益も測っていない。刻みの大きさを一通りに決めており、一分足だけ刻みを宣言で固定しているので他の四つと本数がそろっていない。五つのうち二つは定義から同じもの(ティック足と出来高足)である。本系列の別稿が四つの時計で同じことをしており、本稿が加えるのは五つ目と一枚の表である。予言④は外れた。測定は日中・少数銘柄・四銘柄六年に限られ、月次の時系列モメンタムのようにより長い保有で文書化されている結果を否定するものではない。予測を主張しない。 作成にあたって:本稿の着想と内容は、著者自身の考察に基づくものです。文章の構成整理や英訳、集計に用いたスクリプトの作成には AI(大規模言語モデル)の助力を得ました。最終的な内容の解釈や誤りがあれば、それらはすべて著者の責に帰します。お気づきの点があれば、ご教示いただければ幸いです。 ----- A bar type is a clock. There are bars cut by the clock, bars cut by the event count, and bars cut by the absolute move. So what agrees and what does not when one and the same rule is run on all five? The sentence of the rule does not change by a single character; what changes is only the meaning of "a bar". If the answer changed, then, it was changed not by the rule but by the meaning of a bar. The separator in this paper is what cuts the bar: the clock, the event count, or the absolute move. From ticks for four instruments (USDJPY, EURUSD, GBPUSD, XAUUSD) from 2019 to 2024, five kinds of bar were built: T, minute bars; H, hourly bars; K, tick bars of N ticks each; V, volume bars closing each time volume sums to N; and R, range bars closing each time the path length sums to S. N is 5,120, 4,534, 4,780 and 9,046 ticks and S is 511.6875, 340.5750, 488.2750 and 1749.0050 pips, set so that the counts match the hourly bars. The rule is to enter on the direction of the previous bar (the sign of close minus open) and close 24 bars later; D is the mean of direction times the move over the next 24 bars in pips, with t clustered by day. Cost is the median spread inside that bar, and the spread of a quantity is its ninetieth percentile divided by its tenth. A synthetic random walk (3 draws, times and spreads left as measured) serves as the baseline. Only two things agreed. The direction D lies between -0.0540 and +0.2789 pips and stays below a cost of 0.7123 to 0.7304 pips on all five, and that cost differs by only 1.0254 times from maximum to minimum. Change the clock and what is paid does not change, nor does what comes out. What does not agree, meanwhile, differs by a great deal: the count by 57.4639 times, the duration of a bar by 62.5735 times, and the move per bar by 11.3276 times. The table splits in two, a cost at 1.0254 times against a group at tens of times. The row for D itself is placed under "does not agree", since the sign crosses zero and no ratio is defined; what agrees is that all five stay below the cost, not the value of D, and the two have to be written separately. The diagonal shape — what was held constant is what agrees — appeared as well. The hourly bars' spread of duration is 1.00, the top tenth and the bottom tenth taking the same value, and the tick bars' spread of move is 14.92, the smallest of the five. Conversely the bars cut by the clock spread out in move, at 32.25 for minute bars and 20.63 for hourly bars. Tick bars and volume bars give the same count, D, t and cost, and their edges match at 1.0000, because volume in foreign exchange is the tick count — a check, not a finding. Only the minute bars passed a daily t of 2 (-2.9864), but the size is -0.0540 pips, on a bar whose median move of 0.7250 pips is swallowed whole by a cost of 0.7129 pips. With a count of 2,084,015, even a small size makes t stand up. That t stands up and that the size suffices are two different things. And prediction ④ failed. Tick bars, cut by the event count, give 14.92, below the 15.01 of range bars, cut by the absolute move. The reason is that a range bar holds the path length constant, not the net move per bar: wander up and down inside the bar and the path grows while the net move does not. So the single phrase "the move" was pointing at two different quantities — the path length, which a round trip increases, and the net displacement, which a round trip cancels — and I wrote the prediction without separating them. The same slip of wording appears in another paper of this series. The gap between 14.92 and 15.01 is, however, hardly anything, so what "failed" means here is not that range bars lost but that no gap opened. On the random walk too D runs from -0.1598 to +0.0222 pips and the daily t reaches only 0.1044 in absolute value, the same shape. The conclusion: choosing a clock is not choosing an answer. Choose a clock and the count and the meaning of a bar change, while what comes out does not. What is needed when saying "on this bar type it shows" is not the name of the bar type, but one line putting the size measured on it beside the cost. Scope of this paper: no new mathematical theorem and no new law is claimed; time change and subordinated processes are classical subjects. No real course, author, or vendor is the subject here, and none is named. One rule is fixed and no exit is discussed, the measurement running to a fixed horizon; neither performance nor profit and loss is measured. One step size is fixed, and only the minute bars have their step fixed by declaration, so their count does not match the other four. Two of the five are identical by definition (tick bars and volume bars). Another paper of this series does the same thing on four clocks, and what this paper adds is the fifth and a single table. Prediction ④ failed. The measurement is intraday and covers four instruments over six years, and does not contradict results documented at longer holding periods, such as monthly time-series momentum. No prediction is claimed. Acknowledgement: The ideas and content of this paper stem from the author's own considerations. Assistance from an AI (a large language model) was used for structuring, English translation, and writing the scripts used for the measurement. Any remaining errors or misinterpretations are solely the author's. Feedback and corrections are sincerely appreciated.

View source

Similar papers

#small language model Dataset Open access Oct 2026

Socratic guiding questions in synthetic arithmetic data: matched LoRA runs (revision v2)

Supporting data, adapters, predictions and code for the article *Low-Cost LoRA Fine-Tuning of Small Language Models for Multi-Step Arithmetic Reasoning* by Jake O'Grady, Asena Isik Gürhan, Chee Fong Ting and Effirul Ramlan (University of Galway). We generated 20,000 GSM8K-derived arithmetic problems with step-by-step s...

O'Grady, Jake, Gürhan, Asena Isik, Chee, Fong Ting et al. · 465 citations
#computer vision Open access Jun 2016

Software Development in Startup Companies: The Greenfield Startup Model

The results are packaged in the Greenfield Startup Model (GSM), which explains the priority of startups to release the product as quickly as possible, and the need to shorten time-to-market, by speeding up the development through low-precision engineering activities.

Carmine Giardino, Nicolò Paternoster, M. Unterkalmsteiner et al. · 178 citations · ⚡14
#computer vision Open access Oct 2016

Software Startups - A Research Agenda

Software startup companies develop innovative, software-intensive products within limited timeframes and with few resources, searching for sustainable and scalable business models.

M. Unterkalmsteiner, P. Abrahamsson, Xiaofeng Wang et al. · 157 citations · ⚡17
#machine learning Review Open access Oct 2016

“Failures” to be celebrated: an analysis of major pivots of software startups

This study conducts a case survey study based on the secondary data of the major pivots happened in 49 software startups, and demonstrates that customer need pivot is the most common among all pivot types.

Sohaib Shahid Bajwa, Xiaofeng Wang, Anh Nguyen-Duc et al. · 127 citations · ⚡15
#computer vision Review Open access May 2015

A survey study on major technical barriers affecting the decision to adopt cloud services

The comparison of adopter and non-adopter sample reveals three potential adoption inhibitor, security, data privacy, and portability, which underlines the importance of the technical and security perspectives for research investigating the adoption of technology.

Nattakarn Phaphoom, Xiaofeng Wang, S. Samuel et al. · 111 citations · ⚡8
#computer vision Open access Feb 2018

Lean Internal Startups for Software Product Innovation in Large Companies: Enablers and Inhibitors

This study investigates how Lean internal startup facilitates software product innovation in large companies and identifies its enablers and inhibitors, and shows the potential of the method-in-action framework to investigate the Lean startup approach in non-startup context.

Henry Edison, Nina M. Smørsgård, Xiaofeng Wang et al. · 78 citations · ⚡6

Related blog posts

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.