| StockFetcher Forums · General Discussion · DIVERGENCE IN CORRELATION AS A SIGNAL | << 1 2 >>Post Follow-up |
| guspenskiy13 976 posts msg #119038 - Ignore guspenskiy13 |
4/8/2014 6:32:13 PM I've been looking for an "edge" lately.... and this little concept might give another one for a day-trader like me.. I've had a little scalp on TSLA today and I've been watching it as it bottomed near the previous day close... Everything was telling me to buy the dip...as it turned out to be a V-shaped pattern and ran almost ~5% on a day...but I decided to wait for green... I've ran correlation calculator for TSLA/NDX to find out if there is anything interesting... NDX Green / TSLA Green 55.56% | NDX avg. movement: 0.55% / TSLA avg. movement: 3.11% NDX Green / TSLA Red 44.44% | NDX avg. movement: 0.46% / TSLA avg. movement: 1.54% NDX Red / TSLA Red 80.85% | NDX avg. movement: 0.77% / TSLA avg. movement: 2.6% NDX Red / TSLA Green 19.15% | NDX avg. movement 0.23% / TSLA avg. movement: 3.23% Besides this being a valuable information for any stock you trade IMO, I've found out that divergences between TSLA/NDX in terms of direction and dollar amount can really help your trading. For example, TSLA gapped up ~1.3% in the morning as NDX opened 0.008% green. TSLA filled the gap. When TSLA bottomed and I thougth of a reversal - NDX was down by 0.25%. TSLA was down by 0.44%. Given that TSLA follows NDX ~81% of the time with average movement ~3.376 times higher....that could have been the last confirmation for the decision. On Monday, 4/7/2014 even a better situation happened...NDX gapped down 0.5% while TSLA was down 3%. While I knew that TSLA had good stats on filling the gap - correlation statistics tells me that the gap down was almost ~2 times stronger than the average correlation between two stocks... As the result of this divergence, TSLA exploded and deviation occured on the other side: TSLA was up 1.08% while NDX only 0.15%.... NDX turned red with TSLA being up 0.30%...as soon NDX turned red the odds of TSLA go green to red were 81%...just based on correlation and nothing else... At the end of the day, NDX down 1.11% and TSLA 3.37%... (3.37 avg. strength vs 3.03 end close)....corrected the deviation well.... BTW, the fact that the move was smaller than the average...might give you some insight to the next day price action... P.S. I don't know if anyone would find this useful, but IMO it is a great addition to anyone's trading skills arsenal...not a trading tool by itself, but another insight into price action...an "edge"...I believe.. Especially, if the stocks you trade have big correlation with the market or something else...just watch for deviation... Extremely useful if you trade one stock...and correlation stats are good.... |
| guspenskiy13 976 posts msg #119048 - Ignore guspenskiy13 |
4/9/2014 2:33:27 PM NDX gapped up 0.4% in the morning while TSLA gapped 0.58%. Deviation in the bullish movement strength , went red 2% at 11:00 AM. At 11:00 AM deviation in bearish movement strength in comparison to NDX, 3.8 vs. 3.37, found the bottom and started going up. At 12:10 NDX up 0.7%; TSLA down 1.18%. Gone from the deviation on the positive side to negative extreme and deviation on negative side. If NDX is green 0.7% we are really looking at negative ~2.5% in TSLA. Did not happen, divergence. As of now at 2:30, TSLA is up 0.4%. |
| Eman93 4,750 posts msg #119055 - Ignore Eman93 modified |
4/10/2014 1:20:27 AM Very interesting.. thats where the beta comes in TSLA beta is 1.32 (yahoo) this seems a bit low.. but it could represent an opportunity ... trade the return to to the standard when you have an extreme situation, your example of a 3% gap down in TSLA and NDX down 0.5 so 1.32 * 0.5 = 0.66 if no news or earnings you would have a good probability to return to the 1.32 beta. (mean) I dont know where you get your beta of 3.36.. shorter time frame? I would also use the daily pivot points to help control risk. this is kind of interesting... just found it...read the first few pages... http://www.advisorperspectives.com/newsletters12/pdfs/The_Greatest_Anomaly_in_Finance.pdf |
| guspenskiy13 976 posts msg #119061 - Ignore guspenskiy13 modified |
4/10/2014 2:03:25 PM Thanks for the comment, return to the standard from the extreme - that's exactly what I was thinking. If we have a bearish movement 6 times stronger than NDX, there is a good chance that it will at least return to normal (besides the gap-down stats and other indicators). Similarly on that day, TSLA was up 2%, while NDX faded to 0.1% and eventually turned red. The odds of TSLA to follow NDX on red days during past 6 months higher than 80%; so that was another great opportunity. Again, this shouldn't be used as an indicator to sell - but this is another edge in trading IMO. Considering beta 3.36, when I did it I barely heard of beta. But I had "correlation calculator" that tells me the correlation of TSLA to NDX. And the average movement of NDX to TSLA under those conditions. For example, during the time period of 2 years; the odds of TSLA go green when NDX is green are 58.7%; to go red when NDX is red 66%. I guess yahoo finance uses beta for TSLA since inception, and I believe that is some-what less useful, since TSLA was traded under $40 until April,2013...idk. If I move my time-frame to just 6 months, I get the odds of Green/Green to 55% and Red/Red to 80%. Interesting. The way I got the average movement ratio, in my calculator for certain condition I have avg. NDX move and avg. TSLA. For example, when I use 6 month time-frame Red/Red NDX avg. is 0.78% while TSLA is 2.61%. I just divide one by another and get ~3.346....idk if that's mathematically correct, but it still represents the extremes to the other side... We don't really care about "weak" movements in comparison to the average...just the extremes...better odds IMO. Like when TSLA is green with NDX fading into red...80% chance at least...just from the stats... _____________________________________________________ Considering your article, I found it quite interesting. It actually pictured some-what the approach I've been using lately. I've decided to limit my risk in the stock as much as I could and trade for tiny percentage gains; 0.25%-1%, sometimes a bit more. And increase my position to the max with leverage 6:1. Works very nicely, just trade one stock...it's all you need. People limit their position for "risk management" and they try to find 10-20-50% gainers....maybe it works for some, I prefer the other way around.... _____________________________________________________ There is another observation that I found quite interesting...I've been using ADR (5), as it gives me some kind of sense what the stock can do and what it can't... However, I've been analyzing TSLA more recently...as it's the only stock I trade...and I analyzed ADR (5) and its relationship to the movements...through a week and recorded it, here: THURSDAY: bearish range breakout, but closed right at the range boundary. I believe this movement should continue further. FRIDAY: another bearish range breakout and closed way below the range..I believe there is another bearish day... MONDAY: crazy volatility, but opened with a 3% GAP DOWN. Closed the gap near the HIGH of the range, turned back down, touched the LOWER boundary of the range (complete range fill) and closed higher than the open. I believe bearish movement is over, expect green next day. TUESDAY: GAP-UP open, faded into the yesterday close, then went green on the day, closed below the high. Range fill of 62% in the bullish direction, I expect WEAKNESS next day. WEDNESDAY: weak gap up that went down -2.3% at the peak. Stock closed on green again, but it only filled range for 57%! BIG weakness, expect RED next day. THURSDAY: tiny gap down and TOTAL annihilation in the stock. Currently filled the range at 96% in the bearish direction. Down 4.62% as of now. It is crazy to expect...but according to this logic, we should see another red day tomorrow, at least weak open and maybe some highs into the close. ____________________________________________ I just wanted to show some logic of the concept in my own words, idk if someone would read it. Just based on my weak observation, it seems that it is important how stock acts to its ADR (5)...and we can use it not only as support/resistance.... Some points: 1) Range breakouts in either direction is a strong signal that the stock will head in the same direction at least intra-day for the next day. 2) Direction in which the stock "fills" the range is important. If you compare the open/close and the close is higher (no matter what happened yesterday) and the stock has filled the range - you can see this as a sign of strength in the next day direction. 3) If the fill of the range is relatively weak, like 62% or 57% that I presented, I believe that is a sign of weakness and the movement will reverse. Besides those concepts, there are a bit more, that I am trying to figure out: In the case with TSLA, (1) it either pick the direction right off the open and continues it, until the range is almost filled; (2) it is directionless for 40% of the range and then picks the direction or (3) it fills more than half of the range in one direction and then reverses. My point is...it is hard to explain... but we want to have some kind of statistics.... How much range the stock could "use up" before taking a certain direction... Like...it could be presented in the form of: " 1) if filled range less, than X percent in one direction during a certain time-frame, than likely to reverse" The stock moves in either direction often, at least for first 40-60 minutes...if it fails to fill a certain percentage of range in one direction...then it might turn... I also find that using previous day's close as a signal to buy/sell is very very effective...I've been following TRO's MTC technique - it seems that the previous day's close is a much much stronger indicator... it's the REAL transition red to green or green to red...which is pretty powerful at multiple times... THANKS! |
| guspenskiy13 976 posts msg #119062 - Ignore guspenskiy13 modified |
4/10/2014 2:12:30 PM I just looked at the last concept from the other way - and it makes more sense, if we calculate ADR (5) from the previous close. If we look at the stock, while it kinda goes red, then goes green...we can calculate which side "filled" more range.... For example, if it opens and goes red...then red to green, and just kinda directionless....if it did fill more range on the red side, it might give you better odds to shorting... We use close as the mid-point and watch the movements of the stock in the "red" zone and in the "green zone". Works good on gap-downs as well, TSLA gap down 3 % - still spent more time in the "red" - good shorting on the weakness, while it is in the green zone...As it fails to "fill" the same movement in green, as what it spent in red... Same thing on gap-ups... stock gaps up, kinda volatile in the "green" zone, dips to "red" and quickly leaves...range in "green" is much higher than in "red", good odds to play it long. IDK! lol P.S. Pivot points? I usually use prev. close/open, mid (high-low), sometimes highs and lows when near. How would you use them? P.P.S. Another trade idea that I was thinking about - calculate the average amount of times a "reversal" is happening in a stock, red to green and green to red. This could give you really good odds of achieving a certain percentage gains on some picks. I believe there are stocks that are pretty directional and not too volatile during the day - if they pick the direction, they could run run run... |
| guspenskiy13 976 posts msg #119076 - Ignore guspenskiy13 modified |
4/11/2014 7:40:42 PM Today TSLA was green in a triangular pattern all the way to 13:45, while NDX was down -0.54% at the same time. TSLA had to be at least ~1% down...but it wasn't. I knew the stats...it has been consolidating right near prev.close for 2 hours...while NDX was falling.... Sometimes you just need to be patient...and believe in what you see... 13:50 BOOOOOM = 5% drop. FKIN GORGEOUS I love the MTC concept, but on gap days you can and really should apply the previous close as your signal line.... Because the previous close is the real green to red or red to green transition.... Plus we know the rules how to trade the gaps...we fade them....then trade in the direction, if everything is fine... Using previous close on gap days reinforces this concept.... 5% drop by using simple technical analysis and correlation study with NDX....biggest gain on TSLA to date... P.S. I wonder, whether the relationship of movements between TSLA/NDX at the close can make a difference for trading the next day... TSLA closed -0.20% while NDX is down -1.17%.... I guess this is strength...my ADR (5) concept supports it, as TSLA has used 68% of its range and was pretty much directionless in comparison to yesterday...ADR (5) fill concept tells me tomorrow should be green....Let's see... UPDATE: Since I was trading, I could not compare the gap-down strength of TSLA to NDX. It was -1.7% to -1%. Which suggested, that TSLA is responding to NDX moves less than average. Which also explains why it did not respond to NDX fall mid-day. Strength... |
| klynn55 747 posts msg #119082 - Ignore klynn55 |
4/12/2014 12:04:25 PM some questions: what did you use for a correlator? i know there is a manual one here: http://thedailyvento.blogspot.com/2008/10/how-market-conditions-change-riskreward.html does ARIA fit your bill? i stumbled on it and it produces a 80% ndx red, aria red. |
| guspenskiy13 976 posts msg #119090 - Ignore guspenskiy13 |
4/13/2014 2:38:34 PM I used that calculator, however it's not on the internet anymore, as the web-site of the guy who made it is deleted... I can put it on google drive or something if you want. ARIA (data since 1/3/2014) to RUT (higher correlation): RUT Green / ARIA Green 58.33% average strength: 3.15 times stronger than RUT RUT Green / ARIA Red 41.67% average strength: 6.48 times stronger than RUT RUT Red / ARIA Red 84.85% average strength: 3.17 times stronger than RUT RUT Red / ARIA Green 15.15% average strength: 8.5 times stronger than RUT (Note: correlation on a longer time-frame (1 year) produced 50/50 chance for green Russell and 73.5% for red RUT2K) Personally, I don't see ARIA as a good option for this...I mean, the correlation is strong, but...it's a 1bil company that dropped more than by 5 times last year in couple weeks...anyways.. I looked at last week - the deviation could be played. Look at the chart for Friday ARIA - it almost went green around 1:50 p.m while RUT was tanking in red...good shorting opportunity....7.27 to 7... ~3.5% But personally, I would look at something more liquid with higher m.cap...less opportunities for hype/manipulation...but lower returns too.. |
| klynn55 747 posts msg #119093 - Ignore klynn55 |
4/13/2014 4:00:15 PM i got the calculator from johnny ventos site, its on the calculator, if you look through some of his ventos posts, he tells you how he uses the calculator, question though , you manually find all these using the calculator , or is there a scan on stockfetcher?thanks for your work! |
| guspenskiy13 976 posts msg #119101 - Ignore guspenskiy13 |
4/14/2014 12:39:26 PM Vento uses the calculator to get better odds for shorting stocks. The approach is very effective and I find the calculator very helpful, when I look for shorting opportunities using TRO's HOLO filter. When the market is down and you want to short something from that filter - you have crazy good odds of making 1-3%. You can find HOLO short on this page: http://www.stockfetcher.com/forums/General-Discussion/NEVER-LOSE-AGAIN/67236/940 HOLO long is the next page I believe. ___________________________________ I use the calculator for stocks that I trade every day. Right now, I'm only trading TSLA. I've picked it based on this TRO's MTC checker filter tweaked by olathegolf; it shows the odds of stocks to run from 0.5% to 1-2-3 in both directions. You can also tweak it to get 0.25-0.5-0.75-1 etc for big liquid stocks, like NFLX/TSLA/LNKD etc... Useful for quick scalping and swing trading. you can find this filter on this page: http://www.stockfetcher.com/forums/General-Discussion/NEVER-LOSE-AGAIN/67236/900 ________________________________________ I also use the correlation calculator for the "fade the gap" trades. I always look for an opportunity to play in the direction of the gap, when it's filled, so if the stock has a high correlation with NDX/RUT and the index is going in the same direction as the gap - that gives me also an edge in doing that. You can find many gap filters if you search for fade the gap on SF forum. _________________________________________ Another way to use correlation calculator is when you focus on one stock and you have all the pivot points on the chart, ADR(5), etc.... you actually don't use the correlation directly, but you focus on the deviations of the stock from the statistical correlation....you mainly focus on the deviation the average movement strength and color (red/green)....you want to see extreme situations, such as: 1) stock is green/index is red; while red/red stat is >80 2) deviations at the open with gaps, when average movement strength is higher by 2 and more 3) index is greatly red while the stock fails to follow on the strength (NDX -0.5% while TSLA -0.05%); usually signals a reversal, look at pivots/range/gaps This info is not a trading signal by any means, but it can really give you another edge IMO. If you read my previous posts ^ you might get a better understanding of this concept. _________________________________________ This part is sort of offtopic, but I find it pretty useful and I think this filter wasn't mentioned on SF. This is the filter of JV, the GAP STAT calculator that uses past information to PREDICT gaps....useful tool..change gain/drop_threshold for different results ______________________________________ As I said previously in this thread, always pay attention on ADR(5) and the way it fills; percentage filled, direction filled. It gives you another insight into the stock performance. I also use JV's ADR(5) filter: _____________________________________________ And the last piece....pay attention to pre-market movers. Check them on Gap Stat specific scan for the probability of a gap-fill...correlation could help too... Also - sometimes it is nice to know, if the gap is not filled - how much pct on average it does fill in the case of gap up/down...because there are 100's of situations, when the gap doesn't get filled 100%, but 98-95-90 etc... This link: http://www.financialwisdomforum.org/gummy-stuff/download-stock-prices.htm Excel file that downloads highs/lows/opens/close...basic excel knowledge and you can build a pct statistic for stocks that you constantly play or only play for gaps, so if the gap fills - you know when it is safe to hold and when it's not... Also check this thread for gaps: http://forums.stockfetcher.com/forums/General-Discussion/Realtime-Gap-Scan/99249/10 There are some interesting ideas... _______________ Enjoy. |
| StockFetcher Forums · General Discussion · DIVERGENCE IN CORRELATION AS A SIGNAL | << 1 2 >>Post Follow-up |
Copyright 2022 - Vestyl Software L.L.C.•Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data
Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus