
ïŒEVã§ãç Žç£ããçç±
2017幎é ã ã£ãããªããã®é ã«ãæžãããã©ããŒãžã§ã³ã¢ããããŠåæçš¿
ããŒã«ãŒãæããŠããè¥è ã«ãäœãã®è©±ã®æµãã§ã±ãªãŒåºæºã®è©±ãããã
ãåç60ïŒ ãé åœ2åã®ã²ãŒã ããããã©ãã ãè³ããã°ããïŒã
ââæåãªâç Žç£ç¢ºçâã®è©±ã ã
ãããšæ°æ¥åŸã«ããããªã¡ãŒã«ãæ¥ãã
ãæ¬åœã«æç«ããŠããã§ãããããã°ã©ãã³ã°ã§ã·ãã¥ã¬ãŒã·ã§ã³ããŠãããã®éãã«ãªããŸãããã
æç«ããŠãã®ã¯åœããåã ã
ååãã€ããŠãçè«ã¯ãããççã ããã
ã§ãåã¯æã£ãã
ãçè«ãâçŸå®ã«åçŸãããâç¬éãèªåã®æã§ç¢ºãããã£ãŠããããããšã ãªã
ãããã圌ãããã°ã©ãã³ã°ã«èå³ãæã£ããã£ããã ã£ãã
ð§© ã±ãªãŒåºæºãšã¯
ã±ãªãŒåºæºïŒKelly CriterionïŒã¯ã
è³éãæãå¹ççã«å¢ããããã®ãããçãæããŠãããæ°åŒã ã
åçã pãè² ãã確çã q = 1 - pãé
åœåçã b ãšãããšã
æé©ãªè³ãç ( f^* ) ã¯ãã衚ãããïŒ
ããšãã°åç60ïŒ ã»é åœ2åãªãïŒ
ã€ãŸããè³éã®20ïŒ ãè³ãããã®ãçè«äžã®æé©ç¹ã ã
ð» Pythonã§ç¢ºãããŠã¿ãã
import random
def kelly_simulation(p=0.6, b=1, f=0.2, trials=1000):
bankroll = 1.0
history = [bankroll]
for _ in range(trials):
if random.random() < p:
bankroll *= (1 + b * f)
else:
bankroll *= (1 - f)
history.append(bankroll)
return history
ãã®é¢æ°ã䜿ã£ãŠ1,000åã·ãã¥ã¬ãŒã·ã§ã³ãããšã
f=0.2ïŒ20ïŒ
ïŒã®ãšããæãéãè³éãå¢ããããšãåããã
ååã®0.1ã§ããåã®0.4ã§ãé·æçã«ã¯å£ãã
çè«ã¯ãããã°ã©ã ã§ãåçŸã§ããâçŸå®âãªã®ã ã
ð æåŸ å€102ïŒ (ROI 2%)ã®ã²ãŒã ã§èŠãã±ãªãŒæ²ç·
åç51ïŒ ïŒ= æåŸ å€+2%ïŒã®**ã€ãŒãã³ã²ãŒã ïŒåã£ããåä»ïŒ**ãèãããã
ã€ãŸããè³éã®2ïŒ ãè³ããã®ãæé©ãšããããšã
èµ€ãç·ãæé©ç¹ïŒãã«ã»ã±ãªãŒïŒãç·ããã®ååïŒããŒãã»ã±ãªãŒïŒã ã

èµ€ç·ïŒ2ïŒ ãããïŒãçè«äžã®æå€§æé·ç¹ã§ãããç·ç·ïŒ1ïŒ ãããïŒã§ãã»ãŒåãæé·çãä¿ã¡ãªãããªã¹ã¯ãå€§å¹ ã«äœæžãããŸãã ã€ãŸãã ã±ãªãŒãâååã ã䜿ãâã®ããçŸå®äžçã§ã¯æãå®å®ããŠè³éãå¢ããæ¹æ³ã ããŒã«ãŒã§èšãã°ãçè«äžã®æé©ããå°ãæ§ããã«æã€ããšããç Žç£ãé¿ããæé©æŠç¥ã«ãªããŸãã
å®éã«5ãã¿ãŒã³ãã€ã¢ã³ãã«ã«ãã»ã·ãã¥ã¬ãŒã·ã§ã³ãè¡ããšã
f=1ïŒ ã¯å®å®ããŠäŒžã³ã
f=2ïŒ ïŒã±ãªãŒæé©ïŒã¯æ³¢ãããã€ã€å³è©äžããã
f=4ïŒ ã§ã¯éäžã§æ²ã¿ãããå¹³åã§ã¯åæ»ããã
åãåçã§ããè³ãæ¹ãã§æªæ¥ãå€ããã
æåŸ
å€ãã©ã¹ã®ã²ãŒã ããã£ãŠããã®ã«è³éãå¢ããªãââãããªäžæè°ãªçŸè±¡ãèµ·ããã

ð§® è³ãããããšã©ããªãïŒ
æé·çã®åŒã¯æ¬¡ã®éãã
ãã® ( g(f) ) ã¯ãè³éã®å¹Ÿäœå¹³åçæé·çãæå³ããã
p=0.51 ã®ãšãã®çµæã¯ãããªãïŒ
è³ãç æé·ç ã³ã¡ã³ã 1ïŒ ïŒ0.00005 ãã£ããå¢ããïŒããŒãã»ã±ãªãŒïŒ 2ïŒ ïŒ0.00010 æå€§ïŒãã«ã»ã±ãªãŒæé© 4ïŒ Â±0.00000 ã»ãŒãŒããè³éãå¢ããªã 8ïŒ â0.00020 é·æçã«æžãïŒç Žç£æ¹åïŒ

åã
ã®ãã¹ïŒèµ€ïŒã¯å€§ããäžäžããéäžã§è³éã1/1000以äžã«æ²ãã±ãŒã¹ãå€çºã
å¹³åå€ïŒé»ç·ïŒã¯äžèŠå³è©äžããã«èŠããããããã¯äžéšã®å€§åã¡ãå¹³åãæŒãäžããŠããã ãã
察æ°å¹³åã§ã¯ãããåæ»ãæžå°ããŠããããåãŠãã²ãŒã ã§ãè³ããããã°ç Žç£ãããããšã瀺ãããã
â ïž åãŠãã®ã«è³éãå¢ããªãçç±
åç51ïŒ
ã§ãã49ïŒ
ã¯è² ããã
ãããé¡ã倧ãããªãã»ã©ãè² ãããšãã®æžå°ãææ°çã«å¹ããŠããã
ãããããã©ãã£ãªãã£ã»ãã©ãã°ïŒVolatility DragïŒããšããã
çŽæçã«ã¯ããã ïŒ
â ããŒã«ãŒã§ã®å¿çš
ããŒã«ãŒã«çœ®ãæãããšé£ããã§ããã
åãŠãããŒãã«ã§ããã®è£ã«ã¯è³é管çã®ãªã¹ã¯èª¿æŽãé ããŠããã
GTOçè«ã¯ãåè¡¡ïŒEVã®å¯Ÿç§°ïŒãã远ããã
ã±ãªãŒåºæºã¯ãæé·ïŒlogEVã®æå€§åïŒãã远ãã
äž¡è ãçµ±åãããšïŒ
GTO â çè«äžã®å®ç§ãªãã¬ã€
Kelly â è³éé¢ã®æç¶å¯èœãªæé©å
åã¡æ¹ãšãå¢ããæ¹ã¯éãã
ð¯ è£è¶³è§£èª¬ïŒããŒã«ãŒã«ãããã±ãªãŒã®éç
å®éã®ããŒã«ãŒã§ã¯ãã±ãªãŒåºæºã®ããã«åçŽãªæ°å€ã§ãæé©ãããçããèšç®ããããšã¯ã»ãŒäžå¯èœã
çç±ã¯ã·ã³ãã«ã§âââçã®æåŸ
å€ïŒTrue EVïŒâãèŠããªãããã
ã±ãªãŒçè«ãçŸããæç«ããã®ã¯ã
åç (p)
é åœåç (b)
ãæ£ç¢ºã«ããã£ãŠãããšãã ãã
â ããŒã«ãŒã®çŸå®ã§ã¯âŠ
ããŒã«ãŒã¯äººéãçžæã§ãã
çžæã®æŠç¥ãå€ãã
ããŒãã«æ§æãåã
ã¡ã³ã¿ã«ããã£ã«ããç²åŽãICMã圱é¿ãã
EVãã®ãã®ãæéãšãšãã«æºãã
ã€ãŸããpãããbããåçã§äžç¢ºå®ã
ã ãããããããŒã«ãŒã«ãããã±ãªãŒçè³é管çã¯ãæ°å€ãã§ã¯ãªããæ¹éããšããŠäœ¿ãã
ð åçãšæé©ãããçã®é¢ä¿
åãæåŸ å€ã§ãåœãã確çãå°ãããªãã°ãæé©æè³é¡ãå°ãããªãã

åã+EVã§ããåœããã«ããã²ãŒã ã»ã©ãããçã¯å°ãããªãã
MTTïŒé«é åœåïŒã¯æ éã«ããã£ãã·ã¥ïŒåæ£å°ïŒã¯ããç©æ¥µçã«ââ
ããããããŒã«ãŒã«ãããã±ãªãŒåºæºçãªã¹ã¯ç®¡çãã®èãæ¹ã
ð§ ãŸãšã
åç51ïŒ ã»é åœ2åãªããæé©ãããã¯ãè³éã®2ïŒ ãã
4ïŒ ãè¶ ãããšè³éã¯ã»ãŒå¢ããªãã
ã±ãªãŒåºæºã¯ãEVã§ã¯ãªãã察æ°çæé·ãæå€§åãããçè«ã
ããŒã«ãŒã§ã¯ãè³ãæ¹ïŒãªã¹ã¯ç®¡çïŒãããåã¡æ®ãã®éµã
ãã£ãšæ¥µç«¯ãªäŸ
衚è£ã®ç¢ºçã50ïŒ
ã®ã³ã€ã³ãã¹
衚ãåºããå
šè³ç£ã1.5åã«ãªãã
è£ãåºããå
šè³ç£ã0.6åã«ãªãã
æãéã¯å
šè³ç£ã§ãããæåŸ
å€ã¯å
šè³ç£ã®1.05åã«ãªãã
ããã100åãããš1.05^100ïŒ131.5åãæåŸ
ã§ããã
10000人ã100åãã£ãã

86.4ïŒ
ã®äººã55å以äžã§å
æ¬å²ãã
確çéãã®50å50æã ãšè³ç£ã¯0.0052åã«ãªãã

æåŸ
å€ã¯ãã©ã¹ã§ãããå€ãã®äººãã»ãŒå
šæããã
æåŸ
å€ããã©ã¹ã«ããŠããã®ã¯äžéšã®é£ã³æããæ°åã
BR管çãšããé¢ã§ã¯æåŸ
å€ãã©ã¹ã ãã©ç Žç£ç¢ºçãé«éããã
æ¬åœã«æ°žä¹
ã«ããç¶ãããšãè³ç£ã¯ã»ãŒç¢ºå®ã«ãŒããžåããã