If you know math, statistics, money management, and things that a professional gambler would use to get an edge and manage money, that would help as well.
If you are going to build a trading system, the statistics have to be such that your trades have a positive expectency in the long run, and your overal size of all trades has to be such that you never risk more than you can afford. A winning system can turn to a losing one if you reguarly put a significant amount at risk as your downswings will require greater upside to get back to even. You have to understand that no trading system is a sure thing, but you can calculate the potential profits with a prudent amount of risk per hundred trades by using the Kelly Criterion calculator. Based on this, stocks will hit 20% upside, with 8% downside 90% of the time. IF past results are a decent indicator of future results. We will say only 80% of the time instead with downside of 10%. Now there is a big problem with estimating our wealth increase per trade on average because this assumes we can actually risk 70% of our portfolio. This is impossible because we will only risk 10% if we are wrong. The Kelly Criterion is designed for actual bets. That would mean that we would put up $10 and win $20 if we are right, and lose all of it if we are wrong. That is not the case in stock investing, as we would have to put up $100 in an investment to risk $10 and gain $20. So the maximum bet we really would make is putting up 100% and risking 10% of our bankroll. If you buy OTM options and can figure out your return for the options if you are right, the Kelly Criterion calculator actually will act the correct way when it tells you the percentage you should at the maximum risk, and how much your wealth will increase if you bet that much. Howeverm, you should not bet the full amount anyways. Some may argue that we are really willing to risk a lot more than 10% because the stock could gap down 95% overnight if it’s earnings were all fraud, if it was ruled an illegal business and had to be shut down or something else, but of course the odds of a loss like this are much less, and we don’t actually return twice as much as we lose when we are right. Either way, if we were able to risk 70% (which we are not) our wealth would increase on average about 45.96% per each bet which would consist of a 16 week period. That is not the case, and instead we can expect a 20% return 80% of the time while we expect to lose 10%, 20% of the time. So I believe if we bet 100% we expect an average wealth increase of 14% increase per trade but it may be less due to the fact that when we lose our position size will have to be reduced slightly and a larger gain is needed to make up for a loss.Either way, I think this number is useful in comparing other systems as it has a clear timeframe for the bet as well. I believe this is more favourable than CANSLIM which generally will have the same reward/risk payout, with a lower probability of hitting as cup and handle patterns and double bottoms do not hit their target as often, even though historically the winningest stocks have CANSLIM qualities. However, probably a skillful investor who knows when to hold beyond 8 weeks for winning stocks and knows when general markets are seling off may be able to earn more. I personally prefer consistency that doesn’t require as much skill.
The way we run this system may produce better results than 14% per 16 weeks as indicated above because of the fact that most stocks will be sold between 8 and 16 weeks, some will be sold much higher if they advance more in that time frame. We will allow our winners room to run, by setting stops rather than selling. But it may be less because we will sell some after 16 weeks regardless of how much it’s up and it may take longer otherwise, we do not have the system invest 100% on the first breakout and there will probably be some time when we are in cash between trades as well. And we will not usually be 100% invested. However, those that use margin or options, a tweak can be done to this system so it may produce better results, but that comes with greater risk of course as with 2:1 margin a 10% loss becomes a 20% plus whatever you lose for margin fees/expenses/whatever, plus you are forced out of winning trades if you plan on owning more than one stock. Regardless, even if you only get a 6.5% wealth increase per trade and it takes 16 weeks to get an average of 20% gain on your wins, you still are looking at a portfolio increase of over 20% per year. If you prefer to invest 100% in one flag on the first breakout, you will be 100% invested always except for period between selling and finding another breakout, but you will not be adding to your winners, unless you go on margin, or swap out of your stock for options to add leverage on your 2nd breakout. An option at the money for a stock with 16 weeks or more remaining on the contract (preferably much more like 30 weeks) costs maybe $700 for a call of 100 shares, compared to $10,000. You will probably lose the entire $700 if you “lose” and if your contract expires not up at least $7 you lose some as well. Howeverm if it is up $20 per share, it is worth $20 at expiration which is a huge return. So the return to risk ratio is still roughly $20 to 8 but this time you can add a lot more leverage and your bet is your risk. Additionally, if your stock goes down 8% in a short amonut of time, your option might drop from $7 to $2.5 so when you’re wrong you won’t always lose everything. Also, if you have more time on your contrct than the 16 weeks, selling it at 16 weeks will still give you additional value on the contract and even if it is out of the money after 16 weeks, your downside is slightly more limited. Maximum leverage applied wil in the example allow you to buy as much as 14.2 to 1 leverage. Of course, you should obey the kelly criterion of 70% at risk and you wouldn’t want to own probably even half of this or 35% long options. So you might own $35,000 options at most which would be the equivilent of being long on margin 5:1. In otherwords, if you risk $35,000 you will gain $100,000 if you are right instead of $20,000 if you were long 100% stock, but please do not invest in options without understanding the risks first and paper trading as a large loss can be extremely difficult to overcome. As an example, a 35% loss of your portfolio would require a 54% gain to overcome it. And a loss of 35% of your porfolio twice in a row would require a 137% gain. I feel even 1/2 Kelly Criterion when using options is too aggressive due to so many unknown variables. In theory betting the Kelly Criterion is the amount that if you bet any more you would make less or even lose money as betting any more would result in such large downswings that you would be unable to recover from or at least you would have so much less to bet with that you would be hindering your gains over the long term to the point where it would be more profitable to manage the swings by betting less. Obviously at any point before the kelly, if you bet less you are going to make less as well. However, the closer to the kelly you bet, the higher the risk of ruin. Over an unlimited amount of trades, there is going to be a situation where you lose such a high amount that it is almost impossible to recover from without even greater fortune thanmisfortune. Of course, you aren’t going to be betting an infinite amount, but I feel that for most people a negative string of bad luck will effect them much more negatively than a positive upswing of good luck. Besides how much more than a $20,000 per win in a $100,000 account with a 90% probability of acheiving a win do you need?
What’s more, it’s possible I’m misunderstanding the information. Encyclopedia of stock charts says that there are 54% “throwbacks” in a bull market and 65% throwbacks in a bear market for high and tight flags. Although the average gain is 69% in bull markets and 42% in bull markets, that is not the average price target, that is the average gain from the breakout before it peaks. A throwback I understand is when the stock retraces to it’s initial breakout point. However, the question is will a throwback stop out a stock? As I understand it, the “break even failure” rate is the percentage of stocks showing those patterns that fail to break even. To me, this means that these are the “underperformers” the stocks that don’t advance much. So to me it would seem that the 90% that fail to reach the price target stop out, and if not, the only thing you can really use is the percentage hitting the target and assume that the remaining 10% will stop out or underperform, but it’s entirely possible that the information doesn’t use stops. So perhaps you have to just set a stop at the break even line, and 54% or 65% will stop out, then you have to buy again on a break above that same line again. This will hurt you in commissions, but that would be your only real risk and you could tighten the stop to the breakout point. If this were th ecase the reward/risk may actually be much higher, but you would have to trade it much differently.
——————————————————
A trading system should answer
-When and what to buy
- At what price do you identify buy points
- what price to add shares
- when to sell and take losses
- How to protect gains and let winners run
- When to get out of an underperforming trade
——————————————————
The William O’Niel CANSLIM rules does an amazing job at accomplishing this. However, I think there can be some minor improvements.
1) Only trade the best chart patterns. Cup and handle charts have a higher failure rate and require you to really be able to read the charts not only in the major indicies, but also the individual stocks you are buying. The system can be simplified if you only buy chart patterns which historically prove to be winners even in bear markets. The modification to CANSLIM is that you not wait until a market rally to buy and instead look to buy high and tight flags as stock indicies are oversold in anticipation of a rally while your stocks breakout, potentially emerging as leaders. The preference is to buy CANSLIM stocks that show high tight flag patterns or flags that may not qualify as high and tight flags but are close (up 80% before correcting 25% and then breaking out for example). The next choice is any high and tight flag regardless of growth. The historical correlation is strong enough to show you don’t need fundamentals to produce big winners when there is a high and tight flag. This is the end of the answer to the question when and what to buy.
2) O’Neil has buypoints of high tight flags above new highs. I believe that limits the upside and the other problem is, this doesn’t really allow you a clear cut point on which to add to your winners. If you buy after the stock breaks out of the recent downtrend of a consolidation pattern, you can add higher by buying a second time above new highs. Although ONiel mentions “pyramidding higher” it requires skill to be able to do so. What if the stock rises too fast to add on until it’s too late? That means you miss out on the best stock to add to. The biggest potential winner you will not be able to add to it soon enough, and those that take the longest you will. So the buypoints you identify are BOTH above the recent downtrend as well as at new highs.
3) In case you can’t tell from #2, the spot to add shares is at the 2nd buypoint. This way if the stock fails to march to new highs, you don’t add on, nor should you. This will keep more money in the stocks that deserve it and have proven themselves by rising higher and acting well, while limiting your exposre to those that don’t.
4) There are no changes to how to take losses as a 8% stoploss is still fine… Except that after a stock is up 16% I suggest a 20% or even 25% trailing stop to replace your hard stop, but the purpose of this is more for #5.
5) See 4. A trailing stop will give your stocks room to run to the upside while protecting yuor gains as they continue to march higher. This way as the stock advances more of it’s gains are locked in without requiring you to sell. But ther is more to this. O’Neil puts an exception to the 20% rule. And that is to wait 8 weeks before selling stocks at 20%. The only problem I have here is after 8 weeks if a stock is not up 20%, I’m not really convinced it’s worth owning for too much longer. It’s true you need to sell your stocks up 20% to make up for the stocks that fall 8% and stop out.Additionlly, if a stock is up more than 20% after 8 weeks, it probably is worth holding for longer, or at a minimum worth allowing room to run higher. So I think the following adjustment should be considered….
i) Hold trade until 8 weeks is up first with an 8% stop, then replace it with a trailing stop as the stock advances up 16%.
ii)After 8 weeks, if stock is up over 20% from the 20% profit mark, leave your trailing stop in until you stop out. Else if the stock is up more than 20% but less than 20% above the 20% mark, set a hard stop at a 15-20% profit mark depending on how close it is. I believe a stoploss 5% below whatever the price is is fine.
iii) As if you were making the trade all over again, wait until your stock advances enough so that you can replace your hard stop with a trailing stop of 20-25%
iiv) As if you were making the trade all over again, give the stock another 8 weeks to show it can still perform at a high rate. You are looking to hold onto your performers and get rid of your underperformers. So if the stock can advance 40% in 16 weeks, you can repeat the process. Otherwise follow the rules of #6 to get rid of your underperformers.
6) Although You need to make sure that your gains are large enouh so that the reward to risk is high enough so that your portfolio continues to advance, you also want to get rid of stocks that aren’t performing and get them nto ones that are not. This is why I have a problem with O’Niels rules of always sell at 20% unless the market gives you warnings signs. Since we do not proclaim to be expert stock market analyzers and recognize that that takes a lot of skill, and since that can complicate the whole “take profits at 20% so they outpace the gains” anyways, and since we don’t want to get tock with a stock caught in a -7.9% from your buypoint to 19.99% range, you have to sell. TO me it does make sense to reure that you not sell up 20% unless the stock has been given a full 8 weeks to advance, it seems absolutely backwards and against everything else ONiel teaches to sell those that are up much more than 20% after 8 weeks, while holding on to those that are not. The whole idea is you want to add to your winners and let them run, while avoiding losers. Although a stock that isn’t down more than 8% or up 20% is neither a winner nor a loser, the longer you hold it the more opportunity you miss out on. Therefore, the answer seems to be to follow #5 as mentioned, and to only give a stock another 8 weeks after the 8 week period to hit it’s 20% target, otherwise you will sell it at ANY price.
Additional consirations: Buffett once said it doesn’t make too much sense to diversify too much. Putting money into your 7th best idea rather than more into your first best idea almost has to be a mistake. So I think you should limit yourself to buying at most 5 flags. Your watchlist can be larger than that in case 1 of your flags fail, or another succeeds in breaking out and you still have the ability to add to it really close to the breakout point. But only buy 5 a most and position yourself correctly to do so. It is my belief that you should add more money to the stocks that lead as well as the stocks that are your winners, while putting less into those that lag. So this means the first stock to breakout on your list should have a higher amount of money put in. As soon as that happens, you can reduce the size of all buy orders by half, The same can be said for the first to make it to the 2nd breakout. This way you not only put more money into your winners with a 2nd buypoint, but also to the leaders by adding more to those that breakout first… If one of your best 5 ideas fails, you should also have a slightly smaller position size for your next order. This way as more fail and as the market shows less confidence, the amount of cash you have during this 8 week period increases.
So if you have 100,000 you would asign 20k per stock and set up the buypoints as if you were going to invest close to all of it. As soon as you buy one for 20,000 you reduce yourr buy points by more than half, and set up 2nd buypoints for roughly the same amount. The exact amount can be calculated. You have 1 position you own (assuming only one buypoint goes through) so you have 9 more to go (you will buy the first buypoing for the 4 remaining stocks plus the 2ndbuypoints for all 5 stocks. you have 80,000 to invest in 9 positions, so 80,000 divided by 9 is 8888.88 You probably should make the investments slightly less than this for your first buypoints and slightly more for the 2nd buypoints, so just as you did before, you can be sure you put more money into the leaders. This way once the stock clears it’s second buypoint you would again rebalance your buy orders so that your buy orders are roughly even again. Additionally this has the added bennefit of not putting too much money into the market if the market rally attempt fails and a bear market continues and the stocks you own all fail to reach 2nd buypoint. Finally, if you are forced to sell and clear more cash, you want to look for potential chart patterns to make an additional buy order to your winning stocks, most likely after the 8 week period.
To review:
Identify high and tight flags ( I use a screen of Zacks.com to identify all stocks making a move of 60% or more in the last 12 weeks, above $1 and above 50M market cap. I then start one by one looking through the charts manually starting with the largest advances. Ideally I would like to see stocks that have adanvced more than 80% in 12 weeks, I’m looking for a 100% move from the low or more, a 20% correction or less. I then set up alerts on all of the ones I identify so I know if they breakdown more than 20% from the high. I manually compare all high and tight flags and choose 5 primary or less and 5 backups or less. I look for CANSLIM criteria first, if I can’t find that, I either go with the stock with the best earnings, or generally the largest move from trough to peak before the consolidation as this will result in a higher price target. I also want to see a neat and orderly consolidation, and if the price is high and market cap is large I prefer it as they probably have a lower chance of a false breakout due to volitility.)
Set stop buy orders evenly for 5 or fewer stocks as if you were prepared to invest 100% if every stock brokeout at the same time before you could adjust them. Set alerts for when stocks consolidate more than 20% from it’s highs and if it hits below that mark, you probably want to consider finding a different stock to buy, or at least reducing your position size for that particular stock. Adjust buy orders after the first breakout and corresponding purchase is made out of your 5 stocks so you are set to buy both at the 1st breakout and at the 2nd breakout for all stocks with the 2nd breakout buy orders slightly larger. Then set up stops at less than 8%. One good spot is just below the consolidations lowest low. Another good spot may be 20% below the highs before the consolidation if the stock consolidated near a full 20%. Either way make sure the stop is less than 8% below your purchase. Set alerts for your stocks hitting the price level of 16% or so. It is around this point that you will want to replace your hard stop with a trailing stop of 20-25%. Mark your calender for 8 weeks after your trade and set some sort of alert to check your trade. If you have a price target based upon the chart pattern, once the stock advances above the target price it is acceptable to set a stop at or near 20% before the 8 weeks is up, but it isn’t neccesary. However after the 8 weeks is up, it’s suggested that you give your biggest winners room by leaving a trailing stop to those up 50% or more above initial buypoint, and putting a hard stop around the 15-20% profit marks for all others above 20%. That hard stop will be replaced when stocks are up enough to replace it with a 20-25% trailing stop, and you will give your trade another 8 weeks to show a total of 44% gain from initial buypoint or more. If it does, you will again advance the stop until the stock shows you it cannot advance significantly enough to own. At that point you give it another 8 weeks to either stop out or reach your price target. For all those which have not advanced 20%, give another 8 weeks to hit the price target, at which point you will sell. Set up a limit sell order for the exact 20% price rise from breakout mark.If the stock does not hit the mark within 16 weeks, you will take whatever loss or profit that you have at that point.
http://stocktradinginvestments.com/flag-trading-system-introduction/