Inescapable Truths

You're trying to put words in my mouth. What I originally said was:

You've got it wrong. It's not, "It is not a cat; therefore it is a dog." It's more like, "It is a non-cat; therefore it is not a cat." But that's not exactly what I said either. A4 simply says price does not move in straight lines. How are you not understanding this?

Clearly you are just trying to argue for the sake of arguing. You're trying to make it seem like we disagree, but you and I both know that price usually does not move in straight lines. That's all that A4 says.

I know what axiom is but for market axioms, shouldn't we exclude all those axioms that also apply to random data?

Its true when the positions are forcefully liquidated. People with deep pockets can hold on to negative trades. e.g. Citibank who was just saved from being bankrupt, its shareholders can still make a profit in some (hundred) years. In case of bankruptcy there would have been a final win-loss. So in the end it is indeed a zero-sum game.

Aren't you basically describing a volatility breakout method?

One axiom for markets is fat tail btw and I would definitely add it to the list.

Interesting question. I've been thinking about randomness a lot lately. Some people tend to assume that if prices are random, then it's impossible to be profitable. I'm not so sure this is true...

Also, I think a lot people assume that if prices are random, then the expected change in price is zero. In other words, on average, the up moves will all be cancelled by down moves. This is not necessarily true. Something can be random but still have a positive expected value.

For example, maybe the stock market is random. I don't know, but maybe it is. Let's assume it is for now. Then that would be an example of something random that has a positive expected change in price. How do you know that the expected change in price is positive? Well, take a look at the Dow Jones or the S&P 500 (both approximations for the stock market as a whole). They've been around for many years, and despite all the stock market crashes and bubbles bursting, the price is higher today than it was when the index started. In other words, prices have been generally going up. So there is a positive "drift" to prices of the Dow Jones or S&P, even if those prices are random.

Furthermore, continuing the stock market example, a simple buy-and-hold strategy would be an example of a profitable strategy that works with a random process (again, assuming that the change in stock prices is random). In fact, many portfolio managers believe that the market is random and that you can't beat the market. This means that they believe it's impossible to consistently outperform the S&P 500, for example. So they believe that the best thing you can do is to hold a diversified portfolio of investments. They have profitable portfolios (some of them do at least) and they believe that prices are random.

Of course many people on here, myself included, are hoping for returns much greater than the average return of the S&P 500 (which averages about 12% per year, I think.). So rather than adopting a buy-and-hold strategy, they attempt to beak the market through some sort of analysis.

I believe most methods concocted by Neural Networks/Genetic Algorithms, and the like, surveying a constellation of factors, have a common element. They trade on the basis of what is essentially a return to the normal state. The recursive nature of this phenomenon, the eternal coming and going, keys them in, not Fibs,S/R,etc. Problem:determine the normal state. SubProblem:robalistic bogeymen...fat tails, polarities from multiple means(an expansive process in one data set may be a contractive process in another data set for the same currency at the same time!!, ie.,a breakout away from the mean on the hourly toward the mean on the daily),etc.,etc. The edge provided by any means of trading whatsoever may be ever so slight and only be recognizable over many more trials than what we customarily think of as being a sufficient sample.

I went back and re-read the post, you are correct, I had mis-read your post, I thought you had said 'not straight = waves', so my apologies on that. BTW, I don't argue for the sake of arguing, I'm not tdion, I replied because I thought 4 was rubbish, I'm still not convinced it amounts to anything useful, but I guess we will see.

Price fluctuates/moves in waves/alternates between up & down because of the nature of supply and demand.

To some it may be an irrelevant or unimportant concept, to me it's absolutely vital to success.


The "waves" are there but they don't represent anything in a random candlestick chart, that's the difference. "Waves" formed from price action represent real people making real decisions, and this can be exploited if you are learned enough.

Consider two cases starting at the same point: the middle of a range/consolidating market.

In case A, price thrusts upwards through the top of the range and closes near the high of the bar. The next few bars of case A are low activity down bars with very narrow ranges, before eventually making what appears to be a strong move back towards the top of the range. When it approaches the range, the down move abruptly ends with more strong bullish bars pushing price past the high of the initial breakout bar.

In case B, price thrusts upwards through the top of the range and closes near the high of the bar (same as A). Immediately at the start of the next bar, we have a strong down move back into the range, before it eventually continues the move downwards and breaks out below the bottom range of the bar.

While the situations appeared almost identical, in actuality two very different things were happening. In A, there was strong buying interest that pushed the market past the range and broke out upwards, but there is little follow up either way > no heavy buying or selling. Little interest making moves where price is, so eventually price is pushed back towards the range, and heavy buying returns at discount prices.

In B, what appeared to be buying interest was really just pushing prices higher to get a better price to sell from. When price was at a good price to sell, heavy selling interest entered the market, and just like that what appeared to be a bullish breakout is dead and gone. Not coincidentally, this is when most amateurs would've been buying the breakout. And so goes the nature of the game as the amateur scratches his head wondering why that breakout did not work.

No problem. It happens.

I'm not sure it amounts to anything either. It's just that it's one of the few basic facts about the market that's not directly related to why it's extremely difficult to be profitable (as A1, A2, and A3 are).

Yes, I tend to agree. I don't think the patterns I see on a chart are random; I take them to be patterns in human behaviour, in particular: crowd behaviour.

Also, if you read Mandelbrot's The (mis)Behvaiour of Markets, he explains how randomly generated charts are distinguishable from real charts even though they look the same at first glance (see pages 16-19 of his book). The way that the fake (randomly generated) charts can be identified is if you look at a graph of the consecutive differences in prices, i.e. the change in price from period to period. The random walk (fake) chart is easily identifiable as it has nearly uniform price changes. It's funny, the random walk chart seems to be the least random of them all.

Great discussion, guys!

Yes, this is why I call it "The Golden Axiom."

Excellent post!

I've been wondering if a...let's call it an "anti-chump" method would work. Because of Theorems 1 and 2 (from my first post), I wonder if just doing the opposite of what an amateur would do is profitable. Basically:

 

1. Find a "classic"/"obvious" signal.
2. Do the opposite of what you "should" do.

For example, a "classic breakout" pattern. I think most amateurs would probably enter in the direction of the breakout. Since most amateurs lose, maybe trading in the direction that would return to the range, before it broke out, would be profitable. Indeed, there are many false signals, many "bear traps" and "bull traps."

Mathemetician: I don't think Craig is arguing for the sake of arguing.

He is making the point that wether it's a "wave" or "not a straight line", in order to use it successfully we need to be able to define, measure and quantify. For instance, how many separate waves do we see? Every wiggle? A 10 pip wiggle? A ten pip wiggle within 10 bars?
Not trying to split hairs here. I think Craig is simply suggesting that in order to exploit "waves" we need to start by defining them in a way that could be entered into equations.

anti-chump...you may be on to something. Prices move in waves due to fluctuations in supply and demand. Right...and these fluctuations are orderly,and intelligible, and perhaps form waves themselves???? This is not the way. This is lost in a murky sea. Must find island, a shoreline, a bay in which to anchor. Quantitative methods don't serve one well in this arena. Note the failure GAs and NNs as trading tools. Skeptical empiricism may be better. An island...round numbers. Another...opening range. Another...regression toward the mean. Solid ground. All. Now, make them work.

I like the anti-chump idea. A step in that dierection is to be satisfied with a smaller portion of an anticipated move. The tendency is to jump in quick with a tight stop so we don't miss the move and get a low risk entry. This can cause "Death by 1000 Cuts", or the financial bloodletting of getting stopped out over and over only to see price finally go in the direction of our bias after we've thrown in the towel. Trying to be the first in and last out is an attractive goal but rarely works.

Yes! One thing I've noticed recently about my trades is that I should probably use a wider stop. I would not get stopped out so much, and I would stay with the move longer.

Look at the attached chart. The green trend line comes from the daily chart. It's the main up trend that I'm trying to catch. I entered at point A because price was above the TL and it had recently bounced off the black S&R line. Pretty reasonable. I got stopped out. Then I waited for price to drop a little more and then I placed a new order to get me back in the market right where I got stopped out. This is Jacko's "Anit-Hedging" strategy. It's brilliant if there is a strong trend. It can turn losers into winners. The problem is that I was too tight with my stops.

So anyway, I re-entered at B, then I moved my stop up to just above break-even. (In hindsight it's so obvious that I was being overly cautious because of the loss I just had from trade A.) Continuing on, I got stopped out for a slight profit. I did the same thing. I replaced the order after price had dropped a bit. (This time I didn't wait very long before replacing the order. I was probably looking at the 5-min chart at that point. Again, doing foolish things after getting stopped out.) Luckily, my trade got picked up at point C and price took off. Later I moved the stop up to lock in a bit of profit, but I will wanted to leave a lot of room for retracements because I was targeting something much higher on the daily chart. I got stopped out (again, being too tight with my stops). I re-entered at point D. Finally I got stopped out at point E and price never dropped low enough for me to replace the order. I missed the eventual continuation.

In hindsight, I should have kept my stops wider. I would still be in this trade. There are a lot of chump moves that I must stop doing.

I have a couple thoughts regarding Axiom 2 (that banks and brokers are profitable in the long run), Theorem 2 (that the odds are stacked against me), and what this means for me as an individual trader. Personally, I have no idea what kinds of strategies the banks use. With brokers/dealers, however, they definitely make some of their money (most of it?) through commissions/spreads. They are unique players in this game, and they follow different rules. I assume they speculate a bit, but spreads and commissions are their bread & butter.

I don't know about you, but for a long time I didn't really pay much attention to spreads, especially since they're relatively tight in the forex. However, the spread is the only guaranteed outcome; it has some important implications: For one, I think it makes trading the short time frames extremely difficult. In general, trading shorter time frames involves smaller profit targets (as measured in pips), but generally tighter stops as well. It also usually involves more frequent trading.

When I was really into ultra short-term trading, I remember entertaining ideas where I would have a stop of just a few pips. The initial risk on each trade (stop + spread) was maybe 25% spread, which is absurd! That's extremely inefficient. Then there are those strategies where the potential reward is less than the risk. I've heard a great description of that type of trading: "It's like picking up nickels in front of a bulldozer."

So why should we do the one thing that we know fills their pockets: drive up the commissions and spreads. In general my trading has been moving towards a style that is much more imprecise (i.e. less rigid), more flexible, longer time frame, wider stops, etc. I also remember thinking that I definitely didn't want to trade the longer time frames. I thought that it would be too slow and boring. It's not boring at all; I have something to do and something to look at and check every day. Plus, I'm spending much less on spreads.

1.

The anti-chump method is still usually a chump method. people expect a 'pattern' to 'go in the right direction', set stops (and maybe profit targets), then instead of the market 'doing the right thing' it starts chopping around taking out the chump method as well as the anti-chump method.

2.

the charts on higher timeframes look exactly like the charts of even higher timeframes as well as lower timeframes. charts are simply ambiguous. yes, trading higher timeframes effectively shrinks the relative spread, however stops are then relatively larger.

for a losing strategy it's simply a case of trading a fast timeframe and having to overcome the spread and get stopped out more often with smaller stops, or trading slower timeframes and not having to concern yourself over the spread but instead getting nailed by larger stops. either way a losing strategy is a losing strategy.

if you have a winning strategy and can overcome the spread then a faster timeframe with higher frequency trading will allow a strategy to show itself to still be a winning strategy, or reveal itself to be failing the current conditions... as well as allowing a winning strategy to have the smaller wins compound often. This is as opposed to a longer timeframe where you might have missed plenty of 'wins' on the fast timeframes while sitting on your hands waiting for one bigger play to finish, and in the end still might end-up taking out your stop and kicking you in the nuts.

I simply do not trust ANYONE who trades what i consider to be slower timeframes. As far as I'm concerned they could quite simply be, and probably are, trading a slowwwww lucky streak. By the time they discover all their higher timeframe 'wisdom' has suddenly become irrelevant it could be a few months or a year from now. All that time wasted becoming invested mentally in a strategy that is useless past this lucky streak. Time that could have been spent actually becoming competently profitable.

Overcoming the spread is the biggest hurdle. After that your strategy is either working or it isn't. How long do you want to wait to find out? I could buy stock A or currency B and hold it until i retired and luck-out and become a billionaire... but in the same regard if i'm unlucky i could have lost all the invested money. Such a long timeframe strategy will reveal itself too slowly, and can simply become a case of luck on which side of the sidewalk you end-up living on.

It's a lot like religion. Believe/invest in something you can't know will/won't benefit you until you die, or actually live in THIS life. which makes more sense? ruh roh... don't crucify me plz kthx!

3.

brokers should be regarded like slot machines. yes they will likely have to pay-out someone/s at some point, but all-in-all most people will simply be giving their money to them. Also, even though people walk away with less money than they started, losing money in forex feels a lot like someone giving you the finger... you don't want to take that crap and want to show it who's boss. brokers make money because people keep coming back, even though most apparently lose their wallet each time.

~~~

anyways, those are just my thoughts. not attacking or opposing anyone... it's just how i see things and my perspective. whatever works for y'all is what works for y'all.

Hello nubcake, welcome! First of all, let me say that I understand you're just sharing your thoughts; that's what this is all about. That's why I posted this thread in the "Discussion" section. That being said, it seems you and I disagree on many points. I hope we can have a friendly disagreement. Let me give you my take on your comments:


1. The Anti-Chump Method:

Well this isn't really a method yet. It's just an idea I had yesterday. I haven't done any research on it. All I can say about it is that this was my logic:
A1. It's a zero-sum game. A2. The banks and brokers make a lot of money. Therefore they're taking money away from a group of traders. Let's call that group the chumps. Conclusion: Don't be a chump. Don't do what they do. Do the opposite.
That's about as far as I've thought on that one. So I don't really know if it's a viable method.

2. The Fractal Nature of Charts

I think that's a bit inaccurate. I wouldn't say charts on higher time frames look exactly like charts on lower time frames. I would say they look very similar, but there are differences. One of these differences is that very very short time frames have more zero bars, i.e. bars where price did not change at all (O = H = L = C). Look at the charts I posted. I swear, I just picked two charts that were in front of me. I didn't go searching for charts until I found one that proves my point. You'll notice the abundance of zero bars on the 10-sec and 5-sec charts (EUR/USD and USD/JPY respectively). However, there are no zero bars on the daily chart of the same pairs. Even in the most liquid market in the world, you can't expect a trade every single 5-seconds, but you can expect at least one trade every single day the market is open.

3. The Relative Spread
Looks like we agree on that one. My point is then: Trading at a high frequency will definitely put a lot of money into the broker's pockets. Let's try to not do that. Try and separate myself from the chumps a bit.

4. High Frequency vs. Low Frequency
You are assuming that the same strategy can be applied on any time frame. Well, as I mentioned above, I don't think that all time frames are equal. Yes, I understand about compounding. In fact, that was exactly the motivation I had for wanting to trade very short-term and high-frequency when I first started trading. I would sit there with a calculator and dream up scenarios like, "Wow, all I need it 2% a day and I'll be a millionaire in less than 2 years!" The problem is, I never found a way to get 2% a day. I simply wasn't making money on the shorter time frames. In my opinion, I think it's easier to find a profitable strategy on the longer time frames. Banks and brokers trade short time frames. Why would I want to compete head-to-head with them? (Re)read post #1 on this thread.

I don't let the wider stops of a longer-term trade kick me in the nuts. I adjust all my position sizes so that the initial risk is about 1%-2% of my account.

5. Slow Lucky Streak
If I ever get rich off a really long and slow lucky streak, I'm perfectly fine with that.

6. Overcoming the spread
It's not such a big hurdle if you trade on a longer time frame.

7. Long Term Strategies Take a Long Time to Test
Yes, that's true. However... (1) You could backtest the strategy so you don't have to wait so long. Although I tend to find real-time, forward testing more valuable. (2) You could try several strategies all at once. I could spend a few months testings 3 short-term strategies, maybe I'll give them one month each. Or I could spend a few months testing 3 long-term strategies all at once. Remember, with a long-term strategy, it's possible to follow multiple things at once because it doesn't require your attention every minute.

8. People Give Their Money to Brokers
Yes, exactly. It seems we agree on this one too. The bottom line: Stop giving the brokers your money. Don't drive up the spreads/commissions.

9. A Question For You
Finally, I have a question for you. When's the last time you tried trading a longer-term strategy? Honestly, for me it's just been the past couple months. I used to never even look at the daily chart. I was always focused on the short term, but so far, I like the longer-term trading. And you know what? I always have something to do and something to check every day. So it's not boring at all like I thought it would be before I started doing it.

Great discussion! You made me consider a lot off different things.

yeah man, let's rock this. i just put the 'disclaimer' because some people (probably younger people, who probably aren't even here and don't even know what forex is... assumption++) tend to take my posts the wrong way. Possibly an ego thing that i rub the wrong way and they can't handle. I just do my best to put into words the thoughts in my head, without too much fluff... blunt.

I don't take trading as zero-sum. I suspect brokers can set themselves up to simply be a gateway for you to the/a market, and don't have any risk involved as they don't necessarily need to trade with or against you... or at all. In this case I don't see trading as a zero-sum game... someone is quietly siphoning money as a fee for a service provided, which is fair to me. I guess I'm looking at it from a logistics background for which I have some small experience in... you can have a transport company ship your freight and you pay them, but the fact is they don't necessarily have to actually do the delivery and instead can take a few dollars for themselves and then forward the freight to another service who DO in fact do the actual delivery. The freight company you had contact with were simply providing a service / gateway to allow the freight to get moving, and their part is only small. I see the broker situation as probably being like this in some cases.

Therefore, the broker is simply a third-part who gets to take a dollar here and there for themselves. This doesn't sit well with me as a zero-sum game because in the end the broker can always be in positive cash. I could be wrong *shrug*. The biggest issue though for me is that people regard trading as simply one party sitting on the bid and another party sitting on the ask... and we all know this simply isn't the case. There are other people in the game who sit quietly on the side profiting regardless.

yeah fair call in the way you see them, and I understand what you are saying and seeing and through this view you are correct. Problem is you are being a bit too literal. The thing is that markets aren't truly fractal in reality. The more you zoom-in the more you begin to see anomalies and artifacts, until you get to the tick-level where you get NO chart patterns and only one straight piece of price value vs time. There is no such thing as a open/high/low/close at the tick-level, there is only price values.

Perhaps I'm incorrect, and that is because higher-than tick charts are based on a set timestep, whereas tick-charts are not and only move when price moves, so maybe that's apples vs oranges.

Perhaps a way to look at a normal chart is to consider them all similar and a look at price action through a blurred vision, and the further out you view the charts the more they look the same, but the more you zoom-in the sharper the image becomes until you start to see the oddities that lay within.

So yes, it would be a rare event to see a flat open/high/low/close all of one value on a chart above, say, 15min i guess. The problem as I see it is the further out in time you look the less information you are actually receiving, and the more you zoom-in the sharper the data becomes.

This then comes full-circle to an issue I have with charts in general. Time has been made a component of the data we are looking at, where time may not actually be a relevant component in understanding price action. personally I suspect Point and Figure charts are where the truth lies, but i'm also thinking maybe it's a combination of Point and Figure charts (which give you straight price action without time being any factor) plus normal charts (which DO include time as a factor). Maybe it's neither, or missing another component again... and we are all being lead down the garden path and don't/can't know it. *head scratch*
Yeah, but some things can't be helped. If at the end of the day giving your broker more money also results in you having more money and a superior and more stable strategy... stuff it the broker can have my money then.

In some regards the broker is a bit like having a boss. When you work for someone, you are working to make them money whilst trying to earn a little yourself. People regard trading as a solo business venture where they are the boss and the master of their own universe, but in reality the broker decides how much they are going to dock you for each trade and they are the one's profiting in the end. In an abstract way you are essentially working to make them money.

nah man, i didn't make myself clear. sorry. clearly the spread starts to become a bigger monster needed to be slain the lower timeframe you go... therefore you can't take the same strategy for all timeframes. That's why I said that the spread is the biggest problem to overcome, and which you addressed further down this post. Once you find a strategy that overcomes the spread and is also profitable you then simply flog it for all it's worth... and the shorter the timeframe the more confidence you can hold in the strategy and the more compounding you can take advantage of. it's not about some percentage per day, it's more having a profitable strategy that trades enough so that you have a decent sample size of trades to review to then have confidence in said strategy (and also be raking in a few pips).

over an equal amount of time it's better to have a strategy that is nicely profitable after a 1000 trades than having a nicely profitable strategy after 10 trades. at least, for me i'm going to be liking the 1000 trade strategy more as i'll have 100 times more confidence in it than the 10 trade strategy. stuff the broker taking some of the money along the way, i don't have to like it but it's better to have that confidence in the strategy than to be using a strategy that may give the broker less of your money but gives you less real mathematical confidence (i.e. the sample size over the same period of time is 100 times smaller and therefore hasn't proven itself superior in the long-run to the higher-frequency strategy).

i guess it's like me going on a golf course and my first ever shot is a hole in one... that doesn't make me the best golfer in the world. it means for the small sample size i've proven myself to be 100% awesome, but the sample size isn't worth a thing.

if you can back test and forward test a longer term strategy that makes you money then that is awesome, but if you could have a strategy that is more frequent and still nicely profitable then i'd rather that one (the spread starts to become more of an issue, but if you can overcome that then it's a win) .

hmm, if i understand this correctly... then what you are now facing is a slower frequency strategy that takes longer to play-out, and returns the same amount of $ as you could potentially score if you took the same strategy to a smaller timeframe to the same patterns you have identified as profitable. not sure if that came out how i mean it, but basically if 'winning pattern' in the 1-min chart gives you $1 over 10 minutes... then what i take you are doing is adjusting the risk/reward ratio in a higher timeframe so that you still take away a dollar but have had to take longer to get that dollar, and all for the sake of trying to shrink the effect of the spread.

now on that point, it is my suspicion that there are strategies that people are taking on higher timeframes which are profitable, but which zoomed-in to a certain point will return a higher frequency whilst still being profitable and therefore are optimized for taking the most pips taking the spread monster vs time into consideration. therefore there comes a point where you zoom in too far and the spread starts to become overbearing and you are either earning a less profitable amount over time, or keep going and simply turn a winning strategy into a loser.

basically i think some people trade say 1 hour charts and are taking pips, but if they explored the scary world of smaller timeframes could find a sweet-spot of timeframe vs spread where their frequency over time vs spread rakes you in optimal pips, and therefore at this sweet spot you can take the MOST advantage of compounding.

hell yeah. all the power to y'all that manage to do that. just like the lucky suckers who bought into google on a whim

money is money... whether won or earned. problem is the belief in something that isn't true. others then believe in this fluke strategy and therefore at some point some poor fools are going to get burned. it just takes too damn long for some strategies to play-out before people can realise either way.

see above... already mentioned this part.

yeah, but again the sample size is the deal-breaker in my eyes. it's better in my mind and gut to have a more frequent strategy as it proves itself more and more and you have more time to spread less risk over if you wish, and reap the reward of more compounding opportunity and again a better respect for the strategy as being profitable or not as you find-out faster.

at the end of the day it's about making the most money for oneself. that may, or may not, require giving the brokers a porche or two in the process. if that's what it takes to have a pool filled with money that you swim in then so be it. I think worrying too much (or even at all) about how much the broker is taking away due to your trading is time wasted on something that shouldn't be given any regard to. it doesn't matter what you do they will be taking some money for letting you make a trade, and the only way you can minmize that is to minimize your trading frequency which brings with it the problems i've previously mentioned. To eliminate the broker making money means you can't trade ever. In the end the focus should be on the spread and having a profitable strategy. Thinking about how much the broker is making is mental energy spent on something that should have no time wasted on it. the spread is not what the broker is making... it's simply a slice of the spread that they take. One way or another the spread will be there.

To fight the broker is to fight the wrong beast. Things like slippage etc are the only aspects about the broker that need to be taken into consideration, and like the spread this aspect can be shrunk by less frequent trading... so this is another aspect that needs to be considered if slippage is something that you experience. Again there must be a frequency of trading where the most profit and compounding and belief in the strategy can be realised.

well first things first... i'm not currently trading because i'd be better-off throwing my money into the wind or giving it to a homeless person. i'm trying to create a strategy that is profitable that i am happy with. i'm no expert, and i'm unprofitable. again all i've said have simply been my views, and they should be taken with a grain of salt and people should always remember that my screen name is 'nubcake'.

i don't see myself ever taking a longer term trade honestly. the beast will simply be lying to me in a slower more subtle way, and i'd rather find a strategy that trades as often as it can and continues to prove itself as quick as it can. i'd rather find out i'm wrong quicker than find out over a longer timeframe. when i get it all worked-out i'll be a happy trader. until then there's no need to blow another account. maybe some day i'll have an epiphany and see something about longer term trades that rocks my world, but i will truly be shocked if that day comes. i just don't anticipate this day ever occurring.

so... i wrote somewhat more than i expected there. oops. sorry for the long read

let me know what you think. later y'all.