| StockFetcher Forums · Filter Exchange · A set of trading rules regarding SPY and SH : | << 1 2 3 4 >>Post Follow-up |
| tennisplayer2 210 posts msg #119873 - Ignore tennisplayer2 |
5/15/2014 6:56:15 PM Is there anybody that can make this filter useable for ones with the basic subscription? Thanks. |
| cidrolin 21 posts msg #119881 - Ignore cidrolin |
5/16/2014 7:41:12 AM @gmg733 Well, it's hard to beat a simple buy and hold of SPY during a bull market. I tryed to find entry and exit rule only regarding SPY, and I didn't success. After that I've searched to enter in SPY when ^VIX is decreasing (ie when market is getting more quiet or more bullish I think). One way that poped up was to look at QStick : if Qstick(^VIX,12) < 0 it means that in average, ^VIX is decreasing from open to close for 12 days. **** A simple buy of SPY when ma(5) > ma(200) (from ETFreplay) and sell when ma(5) < ma(200) gives (from ETFreplay) : from 2000 : return of 123 %, max DD of -23 % (SPY buy and hold 76 %, -55 %) from 2004 : return of 90 %, max DD of -23 % (SPY buy and hold 107 %, -55 %) from 2007 : return of 61 %, max DD of -21 % (SPY buy and hold 48 %, -55 %) You can that this very simple approach (moving average) is very effective on a long time, and more important that DD is twice smaller than buy and hold. This week end, 1) I will try to check carefully if my "BullSYSTEM" really beats the simple ma crossover 2) I will try to use my "BullBearSYSTEM" from 2000 (so before the creation of SH) 3) I will try to use Qstick(^VIX,something) > 0 as an exit criterion for BullSYSTEM |
| jackmack 334 posts msg #119884 - Ignore jackmack |
5/16/2014 9:27:48 AM Awesome - I will look forward to your results. Kevin in Ga has done some great work in/on filters like this. There is NO WAY I could touch what he's done in the past with SF and SS - I would love to know what he thinks of this and if he could add to it or help you nail this down :-) Cheers |
| cidrolin 21 posts msg #119917 - Ignore cidrolin modified |
5/17/2014 9:22:12 AM Just to fix ideas (all calculations done with SS) : Buy and hold SPY From 1995 onwards, 9.6 % CAGR, 55 % max DD ; From 2000 onwards, 3.8 % CAGR, 55 % max DD ; From 2007 onwards, 6.2 % CAGR, 55 % max DD ; From 2010 onwards, 15 % CAGR, 19 % max DD ; From 2012 onwards, 21 % CAGR, 10 % max DD ; From 2013 onwards, 25 % CAGR, 6 % max DD ;  hébergement gratuit d'images ma(5) ma(200) crossover bullish : buy SPY From 1995 onwards, 9 % CAGR, 23 % max DD ; From 2000 onwards, 5 % CAGR, 21 % max DD ; From 2007 onwards, 6.9 % CAGR, 21 % max DD ; From 2010 onwards, 9.5 % CAGR, 18 % max DD ; From 2012 onwards, 20 % CAGR, 10 % max DD ; From 2013 onwards, 26 % CAGR, 6 % max DD ;  herbergeur d image Ma(5) ma(200) crossover bearish : buy SH Please note that before 2007, instead of buying SH, I short SPY. From 1995 onwards, -2 % CAGR, 47 % max DD ; Not a good idea !  hébergement gratuit d'images BullSYSTEM : buy SPY Stratasearch entry code: // is bull ? ((mov(symbol(^spx,close),5,Simple) > mov(symbol(^spx,close),200,Simple)) and //entry (adx(8) > pdi(8) or adx(8) > mdi(8)) and (symbol(^VIX, qst(18)) < 0)) Stratasearch exit code: (adx(8) < pdi(8) and adx(8) < mdi(8)) From 1995 onwards, 12 % CAGR, 20 % max DD ; From 2000 onwards, 11 % CAGR, 20 % max DD ; From 2007 onwards, 16 % CAGR, 10 % max DD ; From 2010 onwards, 20 % CAGR, 10 % max DD ; From 2012 onwards, 19 % CAGR, 10 % max DD ; From 2013 onwards, 24 % CAGR, 5 % max DD ;  image a telecharger gratuitement BearSYSTEM : buy SPY or SH Stratasearch entry code: // is bear ? ((mov(symbol(^spx,close),5,Simple) < mov(symbol(^spx,close),200,Simple)) and (cmf(14) < 0)) Stratasearch exit code: (cmf(14) > 0)) Please note that before 2007, instead of buying SH, I short SPY. From 1995 onwards, 8.5 % CAGR, 35 % max DD ; From 2007 onwards, 25 % CAGR, 15 % max DD ; From 2010 onwards, 11 % CAGR, 9 % max DD ; From 2012 onwards, 18 % CAGR, 6 % max DD ; From 2013 onwards, 22 % CAGR, 4.6 % max DD ;  hebergeur image BullBearSYSTEM : buy SPY or SH Stratasearch entry code: // is bear ? ((mov(symbol(^spx,close),5,Simple) < mov(symbol(^spx,close),200,Simple)) and //SPY or SH ? (cmf(14) < 0)) or // is bull ? ((mov(symbol(^spx,close),5,Simple) > mov(symbol(^spx,close),200,Simple)) and //SPY only (mov(close,5,Simple) > mov(close,200,Simple)) and //entry (adx(8) > pdi(8) or adx(8) > mdi(8)) and (symbol(^VIX, qst(18)) < 0)) Stratasearch exit code: // is bear ? ((mov(symbol(^spx,close),5,Simple) < mov(symbol(^spx,close),200,Simple)) and (cmf(14) > 0)) or // is bull ? ((mov(symbol(^spx,close),5,Simple) > mov(symbol(^spx,close),200,Simple)) and (adx(8) < pdi(8) and adx(8) < mdi(8))) Please note that before 2007, instead of buying SH, I short SPY. From 1995 onwards, 18 % CAGR, 32 % max DD ; From 2007 onwards, 43 % CAGR, 15 % max DD ; From 2010 onwards, 34 % CAGR, 9 % max DD ; From 2012 onwards, 20 % CAGR, 9 % max DD ; From 2013 onwards, 24 % CAGR, 5 % max DD ;  photo libre  hebergeur dimage To summarize : 1. bullish crossover on SPY is not so bad ! 2. BearSYSTEM is in par with bullish crossover, with a lower DD and a marginally lower CAGR in very bullish market (from 2012 onwards) 3. BullSYSTEM, BearSYSTEM and BullBearSYSTEM are not efficient before 2007 : with BullBearSYSTEM you will have done significantly worst than the market from 1995 to 2007. 4. If you use BearSYSTEM and bullish crossover on SPY, you would have obtain something very similar to BullBearSYSTEM 5. BullBearSYSTEM has very good stats from 2007 to 2014. |
| guspenskiy13 976 posts msg #119918 - Ignore guspenskiy13 modified |
5/17/2014 12:54:50 PM you should backtest the crossover of (VXX)/(SPY) with its moving average. Can't do that from 1995, but still... |
| tennisplayer2 210 posts msg #119922 - Ignore tennisplayer2 |
5/17/2014 5:24:55 PM Cidrolin, was there a reason that you use ^VIX QST(18) vs ^VIX QST(12)? Thanks for the great work. |
| mahkoh 1,065 posts msg #119927 - Ignore mahkoh |
5/18/2014 7:40:10 PM I made a few modifications to frslbch's spreadsheet to check what would happen if instead of going long SSO on a buy signal one would short SDS, thus using the decay on SDS to your advantage. As SDS has had reverse splits over the period I could not use the next days' open, so I went with the adjusted close of the day entry and exit signals triggered. Where trading SSO turned $1.000 into $ 94.838, shorting SDS ends up with $142.124 over the same period. And there could be squeezed more out of it by avoiding exposure to SDS if the system signals to go short by shorting SSO. |
| guspenskiy13 976 posts msg #119930 - Ignore guspenskiy13 modified |
5/19/2014 12:02:42 AM Why don't you use triple leveraged etfs then? |
| mahkoh 1,065 posts msg #119933 - Ignore mahkoh |
5/19/2014 2:45:24 AM That would indeed further enhance the results, but there are a few practical issues. Historical data doesn't go back beyond August 2009 and SPXU may not have sufficient volume. I think there is a substantial risk of your short being called in on a squeeze. But possibly the strategy can be optimized for ^RUT, in which case one could trade TZA and TNA. |
| guspenskiy13 976 posts msg #119958 - Ignore guspenskiy13 |
5/19/2014 10:18:06 PM true. or TQQQ/SQQQ. |
| StockFetcher Forums · Filter Exchange · A set of trading rules regarding SPY and SH : | << 1 2 3 4 >>Post Follow-up |