Investments Test 2 FIN 4505 Security Analvsis 1 Fundamental Analysis o Most analysts use this o 90 use this o Some information provides advantages o More credibility than technical analysis 2 Technical Analysis o 10 use technical analysis o Does NOT seem to provide that many advantages Moving Average o Technique most widely used for technical analysis o 3 Day Moving Average Day 1 Day 2 Day 3 Day 4 Day 5 Stock Prices 200 201 202 203 201 3DMA 2010123 202 0234 202 0345 o People who use are usually short term oriented o Major Difference information the two use with approaches o Primarily concerned about past stock prices to predict future stock prices 204 203 202 201 200 199 198 3DMA Intersection Day 1 Day 2 Day 3 Day 4 Day 4 5 Day 5 o When the lines cross intersect o Intersection buy and sell signals o Most common type of Moving day Average in real world is 200 day moving average o The intersection on Day 4 5 would be a buy or sell signal 5TOU fteltE f eO ltiNtt A s p i o fill 1l I eo c e 6ft twcd A PI He t r i l1I S ttH V d t o dni c ld i 01 ttJI l o Does Not matter which way the trend is moving only about which line is on top after o Hit and Stock Price is on top Buy Signal o Hit and Stock Price is below Sell Signal o Get signals after the Hit point o Does this work o No evidence proves it is only about 50 accurate o Not so sure it works o No concrete evidence suggests it works o Also used different averages 100 DMA could give you another signal as opposed to the 200 day 50 DMA could give you another signal as opposed to the 200 day and 100 day o If you change the Moving average Change hit points change signal o Is there a reason to use one vs the other o No they are just numbers Efficient Markets negative a b 1 Weak Form Past stock prices cannot be used to predict future stock prices hypothesis in the If it is TRUE you cannot use past stock prices to predict future stock prices If it is TRUE Technical Analysis will not work c Does not affect Fundamental Analysis 2 Semi Strong Form All publicly available information cannot be used to predict future stock prices negative hypothesis a Past stock prices are publicly available information b If true implies that weak form must also be true c If Semi strong true technical analysis will not work and fundamental analysis will not work either because everyone knows information d If Semi Strong is TRUE you will get fair return on the market for the risk you bear e Nothing to do in the markets because it is a waste of time because stocks are all fairly g Semi Strong can be true and still see the market doing crazy things valued f Everyone getting a fair deal SEMI Ta VE J tc i k t off FAU pp E Pl Po lit Ult t t d rlt 1 pri j fS 3 Strong Form All information public or private can not be used to predict future stock prices negative hypothesis a Mergers and Acquisitions good idea of how to this comes into play b Ifthis is TRUE semi strong must also be true pue I c tN t v fee M of M t ft I IbT t EN ioN OfPOt 1 f t f folL W l Ef I ll fh t r S DE J Nf1 LOu If Sf 1 l Ao fY f K c t APPLE 1 PflC s l N II E C TMIS N1 e 4 NtLEIl j C P AI f II fK H l E fMEN f e A flJ Strong form being true extreme case No one literally thinks this is true That is why there is insider trading laws to prevent it because it has been demonstrated that it can be pulled off Empirical Test 1 Random Formula put numbers in a formula and the formula has a statistical test to show if the sequences are random or not i EX 2 4 6 8 10 not random doubling pattern ii EX 1 4 7 10 13 not random 1 If patterns are discernible then they are not random and you could anticipate what will happen next i b Let s put the stock prices for individual companies and the entire aggregate market If formula shows a pattern something we could use to anticipate what will happen in the future If the formula shows no pattern no help to us Ii c The formula detects whether the sequence of numbers has patterns or not d The test has proven that 99 of the time stock prices are random and cannot help us anticipate future prices e Shows that the Weak Form is TRUE past stock prices do not help us predict future stock prices 2 Serial Correlation Test I t 1W It E p l C E ttlI Nc C r t t t l 1 wc c 1 c 1 00 Z 0Cl 1S1 t o c t 1 S 0 1 tlQ I t CiC 1 SI w 1 1 3 1 c t 5 wee CD j ll HN E t PEC1 IOC t Patterns that can Happen Positive Correlation Can use past to predict future If Stock went up Gives you a 50 chance price will go up Same for if the stock goes down Price changes tend to happen in the same direction If this describes reality technical analysis would work C I f I f I Negative Correlation If stock went up stock will be going down next week Same for if a stock went down If this describes reality Technical analysis would work Yes you can predict the future by looking in the past 0 Correlation If stock goes up it s a coin toss 50 chance you don t know what will happen This Correlation is proven to be overwhelmingly found compared to the other two Can Not use past to predict future You could use a regression to be certain of what Correlation your data produces if you do not know what it is 1 positive slope 2 negative slope 3 horizontal slope with no line Also tried to use different forms of t and t l and found the same results tried all the way out to 16 weeks Conclusion is that Weak Form is TRUE 3 Filter Technique 1st step select size of the filter most commonly used is 5 this means stock signals 5 from the high or 5 from the low 101 o Deals with the concept of momentum o Tried filter from to 20 o If small 1 2 you are going to be buying or selling all ofthe time If large 20 mean you will hardly ever trade o o Found out this worked about 50 of the time whiCh means it …