I came across a system recently on Wealth-lab.com that is very similar to the RSI system from VTOReports (http://www.vtoreport.com/rsi.htm) but has an additional feature:
- if the NDX drops 5% from the initial buy, add a double position
- if the NDX drops 10% from the initial buy, add another double position

Theoretically if you buy 1 NDX futures contract, once the NDX drops another 5% you'd buy two more contracts, then two more at a 10% drop, and at the end you're holding 5 contracts. I suppose you could extend it to 15% or even 20%.....however maybe at 15% you'd be better off cutting your losses....and explaining to your wife why you'll both be living out of your car for the next few years.

In any case, the system actually does very well despite most advice I've seen which is to never add to a losing trade. While I can see the logic behind that, at the same time if you have a system which works well doing this I can't think of a reason not to do it. Of course some event could come along and you could get creamed in a crash, but the system didn't do too badly even after 9/11 (you can see the results of the system by clicking on and then entering ^NDX as the symbol, or QQQ, or any Yahoo stock symbol). this link

Since many traders here were trading before I was born I'd be very interested to hear your thoughts on this -- is it a case where conventional wisdom (don't add to losing positions) is wrong, or is the good performance here just a fluke? Personally I tihnk it's not a bad idea, as long as you don't try it with an individual stock. An index sooner or later will bounce even if it's just shorts covering -- if I were short QQQ I think I'd at least lower my exposure somewhat after a 10% move in my favor -- I know I should let my winners run but those short covering rallies during bear markets can be fierce....

SSB

Sorry -- the link ended up at the end of the third paragraph....anyway the "this link" link goes to the web site where you can see the system results by plugging in stock symbols....

SSB

The system will probably work well with stock market indices most of the time.
But I doubt it would work for Notes or commodities.
That is because the stock market tends to be a reversal system.

But the once in a lifetime drop, such as in 1987, can cause a severe drawdown.

Porter

System may work but would you sleep?

Averaging down = no Z's squared

I have two small kids, I don't sleep anyway

However, I see your point. I don't think I could double down unless I was in a very small position to begin with......but the system is good!!!!!

This does get me thinking though -- why not "double down" with a call option instead of actually opening a double position. Especially in a situation where the market is declining so quickly I'd bet calls would be selling pretty cheap....anyway that's something to consider....that also has me thinking that maybe a better way to trade the HiLoLimit system I posted about last week is to buy a call instead of QQQ/SMH/ONEQ --especially with implied volatility near multi-year lows.

SSB

Dollar cost averaging may work better in bull market but not in bear market.

Have you ever tried that system at a casino? It's called the "Martingale system."
That is a system where you double down after each loss until you win.
I tried it once and made several hundred bucks playing blackjack one evening. Then I went home and played it on a computer system. It is a winning system UNTIL you get a string of losses which will wipe you out in no time. That is, by the way, why there are table bet limits. Casino's don't want someone with deep pockets doubling down until they win. In just 10 losing hands with a $5 bet you lose: $5, $10, $20, $40, $80, $160, $320, $640, $1280, $2560 = $5115 in losses just to win $5. But the table limits a $5 table to $500 max bet. That's only seven losers in a row.
I did it once at roulette playing RED or BLACK beginning with a $1 bet. It is very easy to lose $100 beginning with only a $1 bet and doubling down after each loss.
MARTINGALE TRADING SYSTEM

Doubling down with options would only be feasible with far dated options (ie back months) since you would have less gamma sensitivty and the theta decay would be minimized over a much longer duration till expiration. Perhaps the upside to using options would be psychological since you could use the leverage of options to average your cost over a greater range in the underlying. At the same time, I know from experience that once options begin to collapse in conjunction with an adverse move in the underlying, it typically takes a significant move in your favor in a short period of time to recover anything from the purchases. Anyway, try plotting it out paper with options with around 3 months till expiration and let us know how it looks...

Rog,

With all due respect, I totally dissagree. IF you use a "stop," it's not only not very easy to lose $100, it's damn near impossible. If the stop is, say 5 losses in a row, and at that point you go back to your $1 bet, the roulette wheel would have to roll 19 blacks in a row for you to lose $100 (assuming you were betting on red). I suggest that is not "easy". Boring to play that way, but the odds are you will win, not a fortune, but you will win.

1+2+4+8+16=31

Do that four more times and you lose $124. That's 20 losses in a row. It can happen, but it's not easy.

I use averaging in trading futures. I am amused by the conventional wisdom that disdains this methodology. If used with a stop, it works fine for me.

I saw a guy keep adding to a penny stock as it collapsed until the penny stock folded and he was wiped out. He was going to make a fortune, instead he more or less lost his entire trading pile.

Seems to me that people start adding on when they should be stopping out, but some of you are talking some very sophisticated games here.

I wonder if the real bottom line is that there are all sorts of way to skin a cat if the risk/money management is sound.

Averaging UP may be more reliable. Gary Smith, Senor and others seem to do averaging on the buying and/or selling ends. Gary adds purchases slowly as price RISES and comes out fast at 2% declines on his narrow channel upcharts.
Senor seems to sell out in big portions as profit targets are met. Momentum seems to be more reliable but contrarians averaging down are more likely to hit it BIG or lose everything.... Can you catch a falling knife? etc.

Out

40 Bits of Trading Wisdom - Bit #25

Averaging down on a position is like a sinking ship deliberately taking on more water.

I agree with Porter's comments that averaging into positions is really dependent on the markets traded. Stock indicies, in general, will be a bit more "forgiving" as far as scaling into positions assuming that there is a definable range to work with and one is not averaging down or up in the face of a strong trend. Bonds, grains, perhaps currencies can be a much dicier prospect perhaps because the speculators in those markets have a different agenda or the action is more fundamentally driven. To sum it up, I would argue that "averaging down" or "scaling" into positions has to be evaluated on a market specific basis and that markets with bigger mean reversion are the best candidates.

Averaging down is playing with fire. You may get away with it many times, but one day it will clean your clock.

I used to averag down "a little" on futures plays. Most of the time I was hoping to "save" a trade going against me, and sometimes I'd take a loss or 4-5X what I should have on a stop. That's a baaad strategy.

Now, I adhere to a "one price level stop" parameter.

It's much better and safer to stop out and re-enter when you see a new setup.

cap -- the point of the post was to question that wisdom -- not to say it was wrong but to make people think about it. OFten what's known as conventional wisdom is wrong.

As for the Martingale trading system, that relies on having 50/50 odds of winning. You could make the argument that after a 5% decline in the market the odds of a rebound are not 50/50, but higher (due to shorts wanting to lock in profits, whatever). This of course may change the odds.

In roulette each spin of the wheel is completely independent of the prior spin. However in the market today's price (and price change) is somewhat related to yesterday's price and price change so the odds may not be 50/50, and may in fact be in your favor (maybe someone could run the numbers -- or I'll try this weekend). Also since you need an RSI below 30 to even enter the first trade, and then we're assuming another 5% slide, you're talking very oversold and very likely to bounce (unless it crashes, which is very unlikely -- not impossible -- just unlikely).

In short I don't think this system is as dangerous as the Martingale trading system, but that doesn't mean it's not dangerous....

SSB

I have two accounts. One a long term long account that I do not trade. I average down all the time. This is money that I will not touch for 10 years. I think that is more cost dollar averaging. I try to buy the doom and gloom, though. Tell me you hate a sector. That is when I want in. Buying tech in there now

My trading account I use stops. I don't average down but rather get out and then re evaluate my position when things stabilize.

Simple rule: Averaging down is only allowed if it's part of the trade plan, e.g. if you plan to start with a half-position to get a foot in, but expect a pullback and plan to buy the other half lower. That's okay.

But never average down as a reaction to being wrong, that will wipe you out. And as for the Martingale system, the reason it doesn't work is because no one has an infinite supply of capital.

By the way, I do understand the point about the odds of a reversal increasing as you stretch further into oversold (or overbought conditions). From what I've seen people do, it looks like a better way to play that is to wait for the reversal to occur by setting a buy stop (to enter long) or sell stop (to enter short). This helps somewhat to avoid the falling knife. Then once entered, you still have to manage the trade with appropriate exit orders of course.

One more comment: regardless of the odds of success on any given bet, a system where you "double down" is inherently unstable, because the amount of money you have at risk increases exponentially with each loss. Even if the probability of wipe-out is low, if it happens: game over!

Dollar cost averaging is somewhat better because the amount at risk increases only linearly, not exponentially, and I do in fact employ that in my 401k with periodic adjustments based on market conditions.

TS,

You made some great points. I really like the idea of waiting for a reversal before buying back in, that's a great idea. I'm going to see if I can modify the system and only buy after an x% reversal (say 1 or 2%).

SSB

Another thing you'll want to do is use some sort of appropriate trend indicator. Markets that move against trend tend to snap back--a good time to double up. When the trend changes, I say snug your stop and prepare to lick your wounds.

Also, one other time to consider a doubling down strategy is in a severely oversold market. When things get excessive, you know that somewhere along the way, you're pretty darned likely to get a big sharp rally. If you keep buying, chances are, you're going to get bailed out, if only by short-covering.

I'll tell you, though, you want to be patient with that counter-trend approach. Wait for things to get REALLY ugly, and make sure you give yourself enough time to be right.

Mark

Never add to a losing position. A losing position means you were wrong. Jesse Livermore

CenterStageBob,

I think Mark is right on with his trend indicator suggestion. You want the larger, prevailing trend on your side, and if it's not that way, you had best be objectively taken out of the market or pass on the trade.

Lots of good points have been made here. I think you indeed have to control your risk, and if the NDX being down 10% means you're risking 60% of your account, then that's no good. Some would say risk no more than 2%. Personally, I'm a "selective plunger," and I'd go 6% or 8%, sometimes, but also would couple it with a strategy to add to your position as the market moves in your favor.

Not pyramiding, but perhaps something like adding half your original position when the market goes in your favor, twice or three times its normal daily range above your entry, or just picking a percentage as with the downside rules. Add more as the market goes your way.

Let's say we've been in a bull market, and your trend indicators say it's still a bull, though we're getting a pullback. I'd look for chart patterns -- three-wave moves down, bear flags, pennants, "failure swings," etc., and look for Fibonacci retracements as well.

If the NDX has just made a 100 point move up, then for it to go back down 38, 50, or 62 points would be in line with how markets work, quite often, and if we get down 60, for example's sake, and that's where the 5% level is, then great -- a reversal upwards from that area would be expected anyway, per Fibonacci.

The 5% and 10% levels would rarely match up well though, IMO, and I wouldn't go with those percentages, necessarily. I'd go with the Fibonacci retracements in the beginning. If we went up 100, and the market goes down 38 and turns higher, buy (especially if we declined in three waves or if the market looks like it traced out a corrective move down). I'd be willing to add longs at down 50 and even at down 62, as long as our hard science/mathematics trend indicator still says "bull."

If we go as low as down 62 and we're buying, I'd have a hard stop at down 68 or so. Might even buy at down 79, too, with all the above caveats (78.6% is another Fibonacci ratio) with a stop at ~ down 82. Selling the DJIA after a 78.6% rally, compared to a prior decline, has been fantastic since the 2000 top.

I don't know if historically the NDX tends to turn around after declines of 5% and 10%, so I favor the Fibo ratios of prior moves, not percentages of market value. I've seen the ratios work countless times.

Okay, enough.

Best,

Doug

How would it work if you went short from 74/82 low?or long on nsda from 5000???.You need time #1 killer ,,trin and alot of $$$$ can save you.balls help.But;who needs the pressure.

"scaling in" (averaging down) is one thing, "doubling down" is another! If planned from the outset, "scaling in" is fine, but "doubling down" is not. At least that has been my experience.

Contrary... I offer the opinion that conventional wisdom about "not trying to pick tops and bottoms" is wrong.

If you play for tops/bottoms, you at least know quickly whether your trade is probably wrong. Stop out, no averaging against.

"I made 10 times more money trying to pick tops and bottoms than I ever did NOT trying to pick tops and bottoms" --- Paul Tudor Jones

(I too, as a matter of fact)

Hose

Hose, I don't have anything against picking tops and bottoms if it works for you. Waiting for a reversal can increase your odds of success, but at the expense of a good chunk of the price move - it's always a tradeoff! But no matter which way, I agree with you completely about stopping out rather than averaging against. Good trading, TS.

IF you are going to try to pick tops and bottoms, you need to be right on top of the action as best you can.

Many traders want to wait for "confirmation" of the turn. I would argue that there is not such thing as confirmation.

Hose, I agree there is never any guaranteed confirmation. And often when confirmation comes, it's right about the time that the move runs out of steam. But entering with a stop above market (for long) or below market (for short) at least means that there is some momentum on your side (even if it's running out). Either way, you have to be on top of it. I don't think there is any right answer, it's really the age-old debate of momentum vs. value. Whatever works for you... I know that you're a much more experienced trader than I am, so I'm certainly not going to argue against anyone's proven method.

It sounds like your understanding is pretty good, TS.

The most difficult aspect to trading is distilling all of the information, concepts, and beliefs down into something workable.

In the end, the succussful trader must
a. Take sensible risks
b. Try to make SIGNIFICANT gains when given the chance. (Many traders scalp out of a 80 point move after 5-10 points or so.)
c. Exercise rigorous stop discipline.

Sounds easy, but you know it isn't.

Hose

I think some good points were addressed above...

1. scaling in is probably better then doubling down.

2. scaling in is better used at part of the original plan then on top of your full position.

3. even if you scale in there should be some sort of uncle point... you shouldn't be willing to hold til zero.

4. if you commit a survivable percentage to a position you won't get killed... for example, scaling in to a $20K stock position with a $100K account.