股票期货突破技术分析(英文原版)-第29章

按键盘上方向键 ← 或 → 可快速上下翻页，按键盘上的 Enter 键可回到本书目录页，按键盘上方向键 ↑ 可回到本页顶部！
————未阅读完？加入书签已便下次继续阅读！


。1。　
In　a　leptokurtic。distribution；　prices。spend　far　longer　than　they　should　close　to　the　mean。　They　remain　mean…reverting　and　range…bound。　Then；　when　they　
***************　
207　
***************　
It　moves。ahead　of　the　real　world。　It　sometimes。makes　mistakes；　leading　to　the　old　quip　that　the　market　correctly　identified　all　nine　of　the　last　lour　recessions。　
Schumpeter's。model　included　three　cycles。of　different　lengths：　a　very　long　one　that　he　named　after　Kondratieff；　a　ten…year　one　named　after　Juglar；　and　a　four…year　one　named　Kitchin。　
I　use　these　in　the　following　way。　The　long；　Kondratieff　waves。I　call　secular　trends；　as。they　persist　for　longer　than　the　short　cycle。　In　practice　they　tend　to　run　for　multiples。of　the　four…year　wave；　with　sixteen　to　twenty　being　mon　lengths。of　time　to　be　valid。　All　we　need　to　know　at　any　particular　time　is；　are　they　going　up　or　down？　The　sent　of　force　that　drives。these　trends。can　be　technological　innovation；　a　rapid　rate　of　pound　growth　of　GNP　for　a　country；　earnings。growth　for　a　stock；　demographics；　and　expansion　or　contraction　of　valuation。　All　those　forces。themselves。trend。　
When　Great　Britain　was。a　superpower；　the　Kitchin　wave　averaged　4。23　years；　but　it　had　considerable　variation。　An　individual　cycle　could　be　as　short　as　three　years。or　as　long　as。five。　Since　Roosevelt；　however；　the　United　States。has。been　the　superpower；　and　the　cycle　has。bee　phase…locked　to　the　presidential　election　cycle；　which　is。mandated　at　exactly　four　years。　It　is。the　lows。of　the　cycle　that　can　be　locked　down　accurately。　The　analogy　is。with　a　child's。jump　rope。　He　holds　the　ends。of　the　rope　in　his。hands；　but　the　top　of　the　loop　moves。around　rapidly。　In　all　that　time　there　has。been　only　one　time　so　far　when　the　lows。of　the　cycle　were　not　exactly　when　they　should　be。　The　four…year　low　should　have　e　in　1986；　but　the　market　did　not　fall　until　it　crashed　in　late　1987。　The　next　time　that　the　cycle　mistimed　was。in　2006；　as。the　rhythm　indicated　a　market　low；　not　a　high　in　that　year　（see　Figure　10。2）。　

During　the　course　of　a　normal　four…year　period　there　will　tend　always。to　be　a　rotation　of　the　strongest　trends。from　bonds；　to　equities；　to　modities；　always。in　that　order。　Within　the　equity　market　there　will　also　be　a　rotation　of　sectors。starting　with　interest　rate…sensitive　issues；　to　consumption…related；　to　cyclical　laggards。and　resource…related　issues。　This。has　been　well…illustrated　by　the　work。of　Martin　Pring；　Ian　Notley；　and　John　Murphy。　Following　my。system　for　these　rotations。will　lead　to　moderately　active　investment；　but　not　hypoactive。　Typically　I　have　found　that　for　a　large　market　like　the　United　States。or　the　U。K。　the　average　holding　period　for　a　stock。might　be　as。long　as。nine　months。　There　can　also　be　times。when　it　is。nearer　to　three　months。　Total　turnover　is。high　pared　with　a　normal　private　portfolio；　but　low　pared　with　a　hedge　fund。　
Figure　10。3　shows。that　a　secular　downtrend　was。in　place　from　1966　to　1982；　an　era　for　active　management。　From　1982　to　2000　a　secular　uptrend　took。place；　when　a　more　passive　style　worked　well。　Since　2000；　however；　active　management　has。been　required　again。　
Notice　in　Figure　10。4　that；　since　1968；　when　the　S&P　500　index　is。plotted　against　gold　bars。it　has。performed　badly。　What　this。really　emphasizes。is。the　
***************　
210　
***************　

Because　it　is　a　democracy　rather　than　a　munist　state；　India　will　likely　be　the　more　rewarding　stock。market。　Certainly　any　investors。who　ten　years。ago　correctly　identified　the　growth　the　Chinese　economy　was。going　to　have　and　duly　bought　a　China　fund　would　by。now　have　lost　about　70　percent　of　their　money。　India；　conversely；　has。done　extremely　well；　as。Figure　10。6　shows。　

***************　212…213　***************　

tive　trends。　Buys。are　selected　from　the　top　of　the　list　and　sells。from　the　bottom。　There　will　be　a　bunch　in　the　middle　that　are　neutral。　I　do　not　exert　myself　in　trying　to　find　the　very。best　market　or　stock。　That　is。obviously　a　difficult　task。　Instead　I　do　the　easy　thing　and　identify　the　average　or　below　average。　And　I　don't　own　it。　I　define　a　good　stork。as。one　whose　share　price　goes。up　while　I　do　own　it。　The　winners。select　themselves。by　going　up　and　staying　in　the　top　part　of　the　ranking　list。　Simple　though　this。system　is；　it　works。very。well。　I　have　not　found　a　universe　of　markets。or　stocks。on　which　it　does。not　work；　even　though　the　volatility　of　different　universes。varies　widely。　

Trading　Rules　
We　now　need　to　devise　trading　rules。to　get　us。in　and　out　of　our　selected　positions。　I　have　not　invented　any　new　indicators。　I　did　not　invent　the　wheel　either。　It　seemed　to　be　there　when　I　needed　it　and　I　use　the　ones。on　my　car　because　they　work。well。　It　is。the　same　with　indicators。　We　learn　through　back…testing　that　many　indicators。work。well　some　of　the　time。　A　few　work。well　a　lot　of　the　time。　None　work。well　all　of　the　time。　
In　a　leptokurtic。distribution　prices。will　spend　about　a　third　of　the　time　in　a　narrow　range；　and　about　a　third　of　the　time　each　in　an　uptrend　and　a　downtrend。　Each　situation　requires。three　sets。of　indicators。　What　my　method　values。most　highly　is。not　a　range　trade　but　a　strong　absolute　and　relative　trend。　We　use　the　indicators。（like　RSI；　MACD；　OBV；　Bollinger　bands；　and　others）　that　are　set　up　as。standard　defaults。on　the　Bloomberg　Professional…service。　They　are　more　than　adequate　for　our　needs。　There　is。of　course　a　built…in　bark…test　facility。　
Once　the　appropriate　indicators。have　put　us。into　the　positions；　we　simply　hold　on　until　one　of　the　prices。drops。enough　to　hit　a　stop…loss。level。　We　then　sell　out　and　buy　the　next　stock。or　market　that　has。now　been　promoted　up　our　ranking　table。　We　always。remain　invested　until　the　index　itself　gives。a　sell　signal；　by　hitting　its。own　stop…loss。　

This。level　is。a　trailing　percent　behind　the　rising　price。　We　do　not　use　a　fixed　percentage　trail　but　one　that　varies。with　the　market　or　stock。and　is。proportional　to　the　volatility　that　is。normal　for　that　item。　I　call　this。the　volatility。signature；　or　sometimes　I　call　it　beta　with　bells。on。　For　example；　the　number　for　the　FTSE　index　is。7　percent　（see　Figure　10。9）。　In　a　bull　market　this。index　almost　never　has。a　setback。bigger　than　7　percent。　Trailing　up　this。much　below　the　rising　market　index　will　keep　you　in　a　bull　till　virtually　the　end。　However；　a　stock。like　BP　is。inherently　more　volatile；　and　the　signature　works。out　at　13　percent。　The　price　regularly　swings。that　amount　while　still　in　an　uptrend。　The　signature　for　an　index　like　the　BSE　Sensex　（India）　is。18　percent。　
It　is。delusional　to　think　that　we　can　get　away　with　very　tight　stop…losses。and　never　lose　money。　As。Figure　10。10　shows；　if　the　stop　is。too　tight；　you　will　overtrade　and　make　nothing。　If　you　do　not　risk。anything；　you　will　not　make　anything。　We　can　see　that；　over　time；　the　more　volatility　we　can　accept；　the　more　money　we　will　make；　within　reason。　

Of　course；　we　do　have　to　survive　in　the　game。　Figure　10。11　shows。how　much　we　need　to　make　to　win　back　what　might　be　lost　on　a　very　volatile　swing。　If　we　lose；　say；　5　percent；　we　need　to　make　just　over　5　percent　to　get　back。onside。　If　we　lose　10　percent；　we　need　to　make　only　11　percent　to　be　all　square。　If　we　go　down　30　percent；　we　need　43　percent；　a　rather　less　probable　oute；　at　50　percent　we　need　to　double　just　to　get　back。onside。　If　we　were　ever　to　allow　a　drawdown　much　bigger　than　that；　we　would　need　to　shoot　the　lights。out　just　to　get　level。　In　other　words；　we　would　be　eliminated　from　the　game。　In　practice　it　is　worth　tolerating　volatilities。and　the　acpanying　losses。tip　to　25　percent　for　slightly　greater　profits。　Our　stops。are　normally　taken　within　the　range　of　10　percent　to　25　percent。　The　lower　figure　applies。to　big　blue…chip　situations。and　the　higher　one　to　smaller　market　capitalizations；　either　stocks。or　market　indexes。　
The　next　trading　rule　is。to　make　sure　the　size　of　each　investment　unit　is。proportional　to　the　capital　available。　Even　using　the　ranking　system　and　the　correct　stop…loss；　it　is。possible　to　make　wildly　different　amounts。of　money　for　a　given　amount　of　risk。　There　are　essentially。three　ways。to　work。this。　One　is。to　have　equal…size　units。all　of　the　time；　or　unit　position　sizing；　the　second　is。to　use　a　martingale；　and　the　third　and　best　one　is。to　use　an　anti…martingale。　

A　martingale　strategy　is。when　a　man　playing　roulette　always。bets。on　red。　He　knows。red　will　eventually　e　up。　Further；　every　time　he　loses；　he　doubles。his。investment。　When　red　es　up；　he　gets。all　his。losses。back。plus　one。　Unfortunately；　if　the　wheel　has。a　run　against　him；　he　will　soon　find　that　he　is。betting　his。whole　wad　just　for　the　chance　of　getting　a　one…chip　win。　It　simply　is。not　worth　it；　and　he　risks。getting　blown　away。　
An　anti…martingale　is。a　better　system。　It　works。in　reverse。　The　unit　size　is。increased　only　if　the　position　is。winning。　The　unit　is。not　actually　a　constant；　but　a　constant　proportion　of　the　capital　now　available。　If　the　positions。are　losing；　the　unit　size　is。reduced　accordingly。　In　this。way　downturns。are　survived　and　uptrends。are　more　fully　exploited。　Figure　10。12　shows　that；　using　identical　stock。selection　and　entry　and　exit　rules；　a　reasonable　return　can　be　transformed　into　an　extraordinary　performance　with　the　help　of　an　anti…martingale。　
We　can　bine　our　analytics。and　trading　rules。to　build　a　mechanical　trading　system　that　a　puter　expert　would　identify　as。a　rules…based　expert　system。　This。means　we　can　write　algorithms。for　each　part　of　the　system。　There　are　not　that　many　rules。and　they　are　all　simple；　but　their　application　leads　to　a　sophisticated　answer。　They　can　all　be　back…tested；　and　we　can　put　numbers。on　the　probability　of　being　right。　We　have　an　edge；　and　we　know　what　it　is。　There　is。of　course　no　such　thing　as。absolute　certainty。　It　is。more　like　having　a　form　book。at　a　horse　race。　Heavily　backed　odds…on　favorites。do　not　always。win；　but　it　is。the　way　to　bet。　For　example；　the　normal　seasonal　setback　in　September　to　make　a　low　in　October　is。a　70/30　bet。　It　did　not　work。in　2006。　It　did　not　work。in　1986　either。　On　that　occasion　we　had　to　wait　until　late　in　1987　for　a　setback。that　then　took。the　form　of　a　crash。　You　do　not　have　to　be　in　this。business。very　long　before　you　will　be　wrong　about　something。　Nobody　knows。what　the　future　holds。　We　can　speculate；　and　be　right　sometimes。　Our　system　does。not　need　us。to　forecast　the　future；　but　it　does。identify　probabilities。that　we　can　bet　on；　and　provides。a　way　of　exploiting　the　winners。and　minimizing　the　losers。　With　our　system　when　we　are　wrong；　we　can　be　wrong　with　confidence。　
It　pays。off　to　be　in　the　markets。all　the　time。　In　any　given　year　there　are　periods　of　up　to　six　months。when　being　out　maximizes。returns。　This。is。usually　between　May　and　October。　
In　any　four…year　period　there　is。likely　to　be　a　significant　bearish　correction　phase　during　which　swings。down　of　between　20　percent　and　40　percent　are　normal。　Moves。of　50　percent　are　rare；　and　moves。greater　than　that　are　exceptional；　epoch…making；　and　go　straight　into　the　annals。of　legend。　These　drawdowns。do　occur；　however；　and　can　exceed　80　percent。　They　require　subsequent　400　percent　gains　to　get　back。onside；　which　always。take　many。years。　They　are　essentially　terminating　experiences。　
Once　every　ten　years。there　is　a　very　h