Super Stock Picker and risk/money management

[fizikz]

I've been looking at SSP for a while and have been thinking of allocating some funds to a SSP-based Ultimate Price Momentum 4/4+ (UPM4) approach, but what bothers me is that some of the losses are far too big (25-50%) for my taste. Given that the strategy is based on momentum, I would think that it should sell sooner when the price breaks through one or two support levels. Also, I would not feel comfortable buying in the UPM4 strategy when the stock that is picked has begun going down and has broken resistance at the time of the recommendation. I noticed that some of the losing historical trades did just that.

I wonder if others have modified a SSP strategy with stronger risk/money management. I think that the first rule in trading/investing is money management, and that picking stocks is almost inconsequential compared to it. So, even picking stocks using the dart method, combined with strict money management might be a winning strategy. I think SSP does make winning trades at least 50% of the time, so by managing things right, it should be possible to do quite well.

My inclination is to follow SSP's picks but add a few rules:
1. Do not buy a new pick if it has broken a support level
2. On a new buy, place a stop loss order at a support level (and use this risked amount in money management and position sizing decisions, therefore deviating from SSP's weighted allocation rules)
3. After a trade is in a winning position, have a trailing stop at one or two support levels below the current price

What do you think? Suggestions or improvements?

[garilou]

Hi fizikz,

Welcome on the forum!

Suggestions or improvements?

I have absolutely no other link with SSP except for knowing them since very long now, and having discussed with hem and others on this forum a lot.

Si I cannot make any improvement, if you think they are needed, but suggestions, yes!

1. SSP is not only a matter of stock picks, it is a whole trading strategy.

Too many, (including me but for reasons that have nothing to do with the strategy), just "pick" one of the BUY orders.

But before suggesting SSP to change its strategy, you should have learned it very well.

So go through the section
How it works:
http://www.superstockpicker.com/forum/index.php?board=7.0
read all you can, practice on a spreadsheet, and ask questions if you need to, although most questions have finished by being answered with the time.


If you study carefully the SSP strategy, portfolio building and reallocation rules, take the time to fully understand it, (it took me many weeks) - and apply them strictly, you will soon see that, together with the sell signals, even if some losses are part of trading, the portfolios do pretty well, especially UPM4, and there is no need for stop orders.

This does not mean that there are never losses.

2. I just looked at the history of UPM4: the recent big losses ALL happened in a time when the MTI was DOWN, which meant that you bought or held « at your own risks »
I think that you did not take this into account when you wrote your post.

So studying the MTI would be another task.
Because it was created exactly for that reason: signal when to get out of a panicking market.
Again, this does not mean that losses will never occur during MTI UP periods.

You have noticed that SSP gives BUY orders, and SELL orders, but no STOP LOSS orders!
Stop loss orders have been discussed a lot on this forum, and I have studied all stop loss techniques proposed around the Web.

With time, I have come to never place stop orders, even if I have a stop in my own previsions.

I prefer to sell (and often take my profit too early) then wait for a stop that will throw me out either too early or at a lower price then if I had chosen my sell limit.

I agree that if a stock has broken more then one support level, it might well be time to sell. (What really is a support level could be discussed else where.)
 
But you can be pretty sure (even if this stays a part of SSP secret recipe), that it is taken into account.

The fact is that the markets made so many large moves in the recent times, with the MTI changing from UP to DOWN every 2 to 5 days, that it became very difficult to decide to follow or not, to hold or to sell.

Good luck,

Louise

[fizikz]

Hi Louise,

Thanks for your reply.

I realize and greatly appreciate that SSP is not only a stock picker. That would be of limited usefulness. I think the concept of SSP's trading strategy makes sense (momentum), and I am simply trying to have a tighter risk control. I know that the Market Timing indicator was designed for this purpose. My proposal is adding to it.

Also, note that I am not asking SSP to change their strategy. I am simply proposing a modification as an alternative.

I have read a good deal on the forums about SSP's strategy, the MTI, the weighting/reallocation rules (which I find a bit cumbersome and difficult to understand and harder yet to implement). I have made a weighting calculator spreadsheet based on my understanding so far (weight = 1 + perf1 + perf2 + ... + perfn) where "perfx" is the decimalized performance(s) of the stock in the historical holdings. Then, allocation = weight of stock / total weight of portfolio. Almost all of the stocks have a weighting very close to 1. In any case, playing around with these ideas is what prompted me to come up with my modifications.

I know that the MTI prevented a lot of the big losses but it still did not prevent buying stocks in a downtrend. Also, there is the big problem of racking up far too much in commissions in choppy markets like we have now. I quickly checked several historical holdings in UPM4 and my idea of not buying downtrending stocks, putting a stop loss at support, and trailing stops on winning positions would have allowed almost all of the winners to remain in play while cutting many larger losers short without excessive commission costs from entering and exiting too frequently. In my view that is a good improvement.

I noticed that when UMP4 is right, the stocks almost never dip back to the buy price (don't test the previous support), but when they do, it increases the chances significantly that it will be a losing position.

My proposal would be to have a modified UMP4+ where the MTI is not always followed, as a way of minimizing the commission costs of entering and exiting trades frequently in choppy markets. I realize that stop orders are not perfect, but it's the best I've thought of right now. My rationale for using them would be to cut losses quickly and let profits run (even after a sell signal).

To reiterate and elaborate on my modification:

1. Never buy (ignore) a stock pick in a downtrend, or one that has broken support recently.
2. Place a stop loss order at support on a new buy, and only buy when MTI is up.
3. Trail stop orders one or two support levels behind depending on market conditions.
4. When MTI is down, either sell OR tighten stops. This could save on commissions in choppy markets and potentially improve performance by catching sudden upward moves while still having some downside protection.
5. When a SELL signal is received, either sell OR make a very tight stop. This is meant mainly for winning positions and could allow the profits to run while risking a very small amount of the profits.

[DCA]

I have been following '4' for some time now.  I have generally used the MTI for a "don't buy" instead of a sell.
Often SSP picks are of stocks that I already have a position in.  I tend to heavily weight these stocks when SSP indicates a buy.
Sometimes stocks may still go up after SSP sends a sell, but there is one rule that I have learned painfully:  SSP sell means sell (So much that I believe Louise has suggested a short portfolio based on SSP sells.)

Which brings me to a weakness I see in the portfolios.  Sells are delayed by too long.  The SSP sends out sell signals only on a weekly basis.  Often it would be better to jump the gun and send out the sell order earlier in the week.  (On the other hand I have seen stocks go down, SSP not issue a sell order and then the stock rebounds very nicely)

Generally I do not use stop loss orders.  I can usually pay enough attention to the market to take care of it directly.  I do admit that one area I do bend the sell meaning sell is when the sell comes on a stock travelling up.  Then I tend to watch carefully the movement, especially if SSP has a buy that is trending down at the same time.  Certainly if one was going away from the market for a few days a trailing stop sell and/or buy could automate this somewhat.

For an example of why I hate stops in most cases consider recent activity in STB.  This is a sleepy very lightly traded stock with a nice dividend that was blasted by a report from a house that just happened to have just opened a large short position in the stock.  I used this as an opportunity to boost my position at a 30% discount (and a nice one month 13% profit), whereas a trailing stop would have cleared me out along with the other people who scream fire in a crowded theatre.

D

[fizikz]

I did not know that about SSP sell signals being sent only on a weekly basis. That only reinforces my instinctive call to have stricter risk management. I think that my idea of having stops trail positions at support levels is even more justified now.

I was also looking for a way of having short-interest positions, but figured I would look into the long-interest strategy first. I'd be very interested to see a short SSP strategy, that could be used in conjunction with the long strategies. One issue though, is that I am trying to keep the strategies accessible to registered accounts (RRSP, TFSA, etc) and in these cases, the only way to go short is either through inverse ETFs or put options. Many of the stocks picked by SSP are thinly traded and do not have options available, so this makes a short strategy difficult in this case.

I would also be interested to see SSP's strategies applied to US equities, as this would improve the situation with liquidity. TFSAs, for instance, can hold US funds, so this is still a viable investment route for Canadians.

The main trouble with stop losses, in my opinion, is placement. I think the stop loss should be chosen according to the characteristics of the given equity, and based on support levels. Similarly, this is why I suggest using trailing stops based on support levels; simply trailing by a dollar or percentage amount will probably cause unnecessary whipsaw stop-outs.

With regards to STB and similar stocks, it is possible to make good money on their rebounds from drastic moves but in my opinion this is quite risky. Unless I really know the stock very well, I would not want to be exposed to that level of uncertainty and volatility. Part of the risk is due to the price action, but another big risk is in the fact that at that the decisions are discretionary and subjective, and no longer following a defined system like SSP. If I were long in STB, I would probably have sold on Jul 23th or 24th. At that point it looked like the trend was broken.

I think my rule of never buying on the way down will prevent a lot of misery. I prefer to pay a small insurance in the form of missing the initial upwards move in the interest of filtering out false moves. The beauty of trailing stops at support levels is that it keeps you in as long as the trend is intact. If it stops you out, it means the trend is no longer in play and the momentum behind the move is running out. At that point, it's probably safer to find another equity with the wind in its sails than to play with a stock that begins to show too much volatility. I love making profit as must as the next guy, and those large moves can be seductive, but I think that often the significant increase in risk means that it is not the best option. There are probably many other stocks that have a more attractive reward/risk ratio.

[jpenns]

Can you talk more about how you identify support levels?  Are you talking upward trendlines or just a flat pricepoint, or a combination of the two? 

Also, as a stock moves upward, how often do you look to adjust the stops based on new support levels?  I'm asking because it can be difficult to ascertain what's really a support level and what's just a smattering of prices in the same area.

[fizikz]

Good point. For an illustration, I'm referring to something like this when I talk about support or resistance: http://babypips.cachefly.net/school/images/grade1/support-resistance-basics.png

So, by support or resistance I'm implying an imaginary horizontal line defined by previous price action, as viewed on a candlestick chart. For SSP, I usually prefer to use the daily charts.

Trendlines could also be used, but I would look at those more for guessing at retracement levels or profit potential rather than for risk assessment purposes.

[jpenns]

Right, ok that was my original assumption and I've now been studying the charts of recent SSP stocks and trying to figure out if I can apply this into a winning strategy. 

So, next question: your first point is not to buy stocks in a downtrend.  How do you define downtrend, is that a longterm trend, or would a week trend suffice?  How would you view the purchase of DOL on May 22?  At that point there's been a week's downtrend followed by a few days of sitting at a support level.  Would the strategy be to buy it unless it sank below that support level?

In that particular case the point is moot because the MTI was down, so your strategy wouldn't buy the stock until June 18.  A big run would be missed, and by June 18 the price is up to 62.50.  Following your strategy I would choose not to sell on July 23 when the sell signal came, but instead set a stop at 63.00 where there is recent support.  That stop would be triggered a few days later, and my profit on DOL would be less than 1%, whereas SSP posted a gain of 16%.  Is that just the price of having smaller risk involved?  Have you done a study on your strategy to see if the mitigated risk is balanced out by also cutting out some big wins like DOL?  So far I've gone back as far as DOL, and the results have been mixed.  Some small increased gains on other stocks, but huge missed opportunity on DOL.  I suppose a telling test will be seeing what the difference is with some of SSP's big losers.

I'm not trying to tear your strategy apart, I really want it to work, I just want to see from past performance that it could work before I commit my money to it.

[fizikz]

Firstly, I would not suggest putting money on this strategy yet. I would want to have rules for every possible situation, and backtest them, before encouraging others to try it out. The purpose of this discussion is to come up with ideas on how to implement tighter risk control with SSP. The MTI was created in that spirit, but I think we can improve on that for those who have a lower risk tolerance.

About not buying in a downtrend, I mean that both from a longer term and shorter term point of view. In a longer term I think it's unlikely for SSP to pick downtrending stocks, so that point is moot. In a shorter term, I don't want to be buying if it looks like the stock is going into a retracement, or maybe if the uptrend is over. If it is really a retracement, I prefer it to reach support, test it, and then buy on the way up. If it never finds support, or if the trend has changed to the downside, then I avoid or minimize the losses by not buying on the way down.

Specifically for DOL, I see a smattering of flat prices between April 24 and May 4, and the low set there was 53.50. When the buy signal came in on May 22, the stock had gone through a retracement, but the last two candles were green indicating that perhaps the stock could be continuing on the uptrend. Also, the low in that retracement was 53.51. It did not break the support, and it looked like May 22 was a perfect entry point to continue with the uptrend. I'll grant you, though, that it this is being said in hindsight.

One point that I'd like to make is that often the support and resistance areas are not exact, but more of a region. Considering that SSP's strategy is designed for a longer term and not for day trading, I think it needs to be given more breathing room, which is why I was saying to place stop losses at one or two support levels below the price depending on how tight you want to be with risk in view of market conditions. I think that being too tight with stop losses would be problematic with SSP.

The other point you bring up about the use of the MTI is also important. I know I mentioned not buying when the MTI is down, but I'm not entirely convinced of this. The MTI is designed to get a sense of the market direction, no doubt using some moving averages. However, in any market, some stocks are going up and some are going down. In a downtrending market there will still be many stocks going up. It might just mean we need to be more careful with the downside risk. In principle, I would prefer to trade in the direction (up or down) of the overall market, but trading the downside is not easy with SSP or in registered accounts. Since we are trying to manage the risk for each individual trade based on its own support levels, the idea of following an indicator for the whole market becomes less important for the individual trades, in my opinion, although I would not ignore it completely. I'm trying to see what kind of indicators (moving averages, parabolic SAR, etc or combinations) might work well for allowing entries, and also for determining exits.

The bottom line is that the modifications are meant to tailor the risk management to each trade instead of using the whole market to determine moves to be made on every stock held, or to be held, in the portfolio.