Grid trading: Does it work?

What is grid trading?
And why is it frowned upon? (if other sites forums (fora?)) are to go by)
Does it work? If not, what is conceptually wrong with it?

It does not.If price makes a large move in one direction it will lose a load of bucks.

http://fxtrade.oanda.com/community/fxmessage/ultimatebb.cgi?ubb=get_topic;f=15;t=003696;p=1

Right, so it works, then. Hey, I'm just kidding, ODT. Am I right in thinking this is a form of averaging in?

Dommo said:

Right, so it works, then. Hey, I'm just kidding, ODT. Am I right in thinking this is a form of averaging in?

Click to expand...

No.It is no stop loss and keep adding to losing positions. .

You haven't told him why it will not work!

There have been posters who have said that they make fortunes out of it.

I have tried all sorts of ideas but, frankly, I am absolutely too frightened to try that one.

i thought it was calculated averaging down with small size

Splitlink said:

You haven't told him why it will not work!

There have been posters who have said that they make fortunes out of it.

I have tried all sorts of ideas but, frankly, I am absolutely too frightened to try that one.

Click to expand...

He can't be that mad not to understand what I never told him.

oildaytrader said:

He can't be that mad not to understand what I never told him.

Click to expand...

ODT, to me it looks like a form of averaging in (which you could formulate in many ways); is there some critical difference? Cheers

yeh it's averaging down with small size to begin with

I am going to make grid-trading workable!

If you're averaging down continually, you're using a martingale system, and sooner or later you'll lose all that you risked. In the meantime, you'll win every trade and your account will grow.

How skillfully you add positions will determine how soon or late that blow up point is. You should be able to last for quite some time

If you start with an initial account, and use a grid system and withdraw all profits, it is possible that by the time you did reach that blow up point, you would have already doubled, tripled, quadrupled intial account. You can then start again, rinse and repeat. As long as you have the discipline to take that final loss and not risk your profits

Shakone said:

If you're averaging down continually, you're using a martingale system, and sooner or later you'll lose all that you risked. In the meantime, you'll win every trade and your account will grow.

How skillfully you add positions will determine how soon or late that blow up point is. You should be able to last for quite some time

If you start with an initial account, and use a grid system and withdraw all profits, it is possible that by the time you did reach that blow up point, you would have already doubled, tripled, quadrupled intial account. You can then start again, rinse and repeat. As long as you have the discipline to take that final loss and not risk your profits

Click to expand...

Trading griding is loser's mindset, it affects other trading and breeds bad discipline.The last persons did not have discipline to lock away profits, but put profits back toconverting them to losses.

If he gets hit before he makes any money , he will be O K

The idea of grid-trading is to keep multiple, and increasing numbers of positions open.
The increasing number of gains of closed trades tend to to be offset by an equally increasing amount of liability in the open trades.
The practicality of paying for overnight charges or fees, or interest, makes this idea potentially nett losing over time. Not to mention the margin issues for the opening positions.

The underlying principle is one of mean reversion.
Two accounts are needed. One for the accumulating buys, another for the accumulating sells.
"You should try to maintain a positive interest carry, if possible." (comment from the thread ODT referred to)

But, I am sure there is a way around this...

trendie said:

What is grid trading?
And why is it frowned upon? (if other sites forums (fora?)) are to go by)
Does it work? If not, what is conceptually wrong with it?

Click to expand...

Probably not.

trendie said:

The idea of grid-trading is to keep multiple, and increasing numbers of positions open.
The increasing number of gains of closed trades tend to to be offset by an equally increasing amount of liability in the open trades.
The practicality of paying for overnight charges or fees, or interest, makes this idea potentially nett losing over time. Not to mention the margin issues for the opening positions.

The underlying principle is one of mean reversion.
Two accounts are needed. One for the accumulating buys, another for the accumulating sells.
"You should try to maintain a positive interest carry, if possible." (comment from the thread ODT referred to)

But, I am sure there is a way around this...

Click to expand...

Definately not.

Shakone said:

If you're averaging down continually, you're using a martingale system, and sooner or later you'll lose all that you risked. In the meantime, you'll win every trade and your account will grow.

How skillfully you add positions will determine how soon or late that blow up point is. You should be able to last for quite some time

If you start with an initial account, and use a grid system and withdraw all profits, it is possible that by the time you did reach that blow up point, you would have already doubled, tripled, quadrupled intial account. You can then start again, rinse and repeat. As long as you have the discipline to take that final loss and not risk your profits

Click to expand...

Don't forget to include commissions and the like...

Can you make money? Yes

Do you have to be lucky? Yes

im confused at this .... thought you meant averaging in to trades which makes sense but this sounds a bit bleurgh

fwiw my general opinion on this is that is:

* useful if you are a market maker / liquidity provider / pure scalper

* bad at any other time

The problem, as I see it, is the small linear accumulation of profits, but the geometric accumulation of liabilities.

for example, if the grid was:
1.7
1.6
1.5
1.4
1.3

a:if you have a buy/sell at 1.5.
b: then market moves to 1.6.
(at this point, you close out the long for +100. but, immediately open another buy/sell.
you have encashed 100 pips, but have 2 open sells, and new open buy. net gain zero)
c: market moves back down to 1.5.
(you now close out the sell (opened at 1.6) for +100. theoretically, you should open another buy/sell.
but, you now have +200 pips in the bank, an open buy at 1.6 (-100), nett 100 pips.

but....
a: if the market moves from 1.5 to 1.6
(you cash the 100 pips, and open another buy/sell)
b: market moves to 1.7
(you cash another 100 pips from the buy, an open another buy/sell.
you have cashed out 200 pips, but have a sell from 1.5 (-200) and a sell from 1.6 (-100), nett exposure -300)
c: market moves to 1.8.
(you cash the buy for another 100 (nett banked 300 pips), but are required to open another buy/sell.
you have liabilities of -600! -300 from 1.5, -200 from 1.6 and another -100 from 1.7.)

The nett result is a linear growth, but geometric liability.
However, grids are recognised as being based on the assumption of reversion to mean.
That is, money is made by riding out stomach-churning losses until the reversion kicks in. It is acknowledged that a strong trend will eventually break the grid. But, even if the grid reverts, depending on the time it takes to revert, the "notional" return to zero is at a cost of interest payments, etc, while the trade has been underwater. So, in truth, it is actually a small loss.
However, if the grid spacings are wide enough, you can ride out a bigger trend. But these trades will logically last longer, and the cost of the trade is expensive as you have to pay to keep it open longer while it returns.

I am still convinced there is a way out of this. I am determined to find a viable mechanism.
One note is that there appear to be some variations on the grid system.

Since the principle is based on not closing out a position, and essentially to aggregate them over time, why not adapt the principle of "not closing out a position" to any indicator based system?

For example, any old MA-X-over system could be used to trigger buys and sells. You close out any winning position to bank profits, and leave any losing positions open until they return to median to close out,
OR, close out an open long and sell when their aggregate is zero?? (eg, if a losing buy is -620, close out this when an open sell can be matched against it when the price reaches -310, an aggregate of zero.) This might reduce time of open losing positions and the resultant trading costs, as well as free up margin more quickly.
This way a bias is introduced into the trading.
Additionally, in order to avoid positions to be opened too close together, ie, when market is in chop-mode, you refuse any new trade unless it is greater than a defined "grid" size. eg, if the market is choppy, and you open a trade, either buy or sell, and you have a grid size of 100, you refuse any new buy/sells unless they are at least 100 pips away. This avoids collecting a heavy position at a narrow price range.
This is a possible variation on grid-trading; just leave positions open.
The advantages are the addition of a bias, the principle of leaving positions open as per grids.
Being a trend-follower, my mindset is to wait for the one-directional trend, which is the money-maker, whereas grids fear the trend, as it kills them in one single shot.