when looking at 5m,15m and 1 hr, how long does each usually cycle? i.e. in the 1m, I noticed that peaks and valleys tend to form about every 3hrs, assuming no news reports or aliens are attacking
Over the long haul, you will find “usually” to be rather unreliable.
Line up those swings you see with the different markets opening and closing throughout the day (Asian, European, and US) and you might find a bit more reliable correlation.
Ya I wouldn’t try to figure out how often they occur and just realize that they do.
There are no patterns in those time frames, they are purely random.
Care to elaborate more on that?
…the M5, M15 and H1 have “no patterns” and are “purely random”?
Just confused.
Thanks!
Random meaning without a pattern. One can profitably trade chaos, but that does not make it patterned.
I disagree.
A truly profitable setup is robust, and fractal in nature.
Time frames with peaks and valleys…sorry, don’t think the time frames have anything to do with it…if it were that simple…smh
Find the pattern:
On Monday,
Joe sells 1 pair of Nike Superduperdopeness kicks at 3:00pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 3:54pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 5:32pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 6:17pm.
On Tuesday,
Joe sells 1 pair of Nike Superduperdopeness kicks at 4:47pm.
Joe sells 2 pairs of Nike Superduperdopeness kicks at 7:31pm.
On Wednesday,
Joe sells 1 pair of Nike Superduperdopeness kicks at 2:01pm.
Joe takes a return on 1 pair of Nike Superduperdopeness kicks at 3.58pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 6:42pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 6:49pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 7:18pm.
On Thursday,
Joe sells no Nike Superduperdopeness kicks even though he sits in the store hoping to do so all day.
On Friday,
Joe sells 1 pair of Nike Superduperdopeness kicks at 4:07pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 4:39pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 5:32pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 6:46pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 7.13pm.
Joe sells 1 pair of Nike Superduperdopeness kicks at 7:43pm.
The pattern? Joe is trying to sell Nike Superduperdopeness kicks Monday through Friday.
The number of pairs Joe sells each day and the timing of each sale makes no pattern. The data set is the result of random uncoordinated events.
There is a trend on these shoes. They have outsold all other models in the store this season. The buyer has restocked this SKU three times in the last 6 weeks. But there is not a pattern that if understood can enable any prediction of the time at which a customer will walk in and buy a pair of these shoes. Even if one perceives that there is a pattern in the times and numbers of pairs sold, it is simply a coincidence made from pure randomness. On the subject of apparent patterns from randomness: Fooled By Randomness - Nassim Taleb
Joe is still profiting from the random transactions that occur throughout the day and he is not the least bit concerned that his profits come to him randomly. His profiteering does not have anything to do with the perception of patterns in the times and numbers of pairs sold at each subsequent sale.
The popularity of the shoe model is a trend. Buying the shoe is popular and Joe benefits from that popularity. But he cannot do so by finding any pattern within the times and numbers of pairs within the data set of all the sales for the day. The social bias prevailing in the society to buy the shoes can only be quantified in longer terms in aggregate values (the SKU has sold more pairs than any other this season).
I would agree with that…
if we were selling shoes…
Find the pattern:
On Monday,
Trader Joe sells 1 microlot of USD/JPY at 3:00pm.
Trader Joe sells 1 microlot of USD/JPY at 3:54pm.
Trader Joe sells 1 microlot of USD/JPY at 5:32pm.
Trader Joe sells 1 microlot of USD/JPY at 6:17pm.
On Tuesday,
Trader Joe sells 1 microlot of USD/JPY at 4:47pm.
Trader Joe sells 2 microlots of USD/JPY at 7:31pm.
On Wednesday,
Trader Joe sells 1 microlot of USD/JPY at 2:01pm.
Trader Joe covers 1 microlot short of USD/JPY at 3.58pm.
Trader Joe sells 1 microlot of USD/JPY at 6:42pm.
Trader Joe sells 1 microlot of USD/JPY at 6:49pm.
Trader Joe sells 1 microlot of USD/JPY at 7:18pm.
On Thursday,
Trader Joe sells no microlots of USD/JPY because he did not get a signal.
On Friday,
Trader Joe sells 1 microlot of USD/JPY at 4:07pm.
Trader Joe sells 1 microlot of USD/JPY at 4:39pm.
Trader Joe sells 1 microlot of USD/JPY at 5:32pm.
Trader Joe sells 1 microlot of USD/JPY at 6:46pm.
Trader Joe sells 1 microlot of USD/JPY at 7.13pm.
Trader Joe sells 1 microlot of USD/JPY at 7:43pm.
Joe likes to sell.
Exactly. The major U.S. firm, having posted a €100,000,000 to their receivables upon invoicing a big deal in the eurozone, that calls up a broker and puts on a €100,000,000 hedge position in the forex at 5:23 did not do so according to any pattern. The few pips the EUR/USD moved as that position was taken on are not part of any coordinated series of events. In fact, the accountant at the firm made the call after a long unscheduled bathroom break which put the call off 17 minutes.
The movements throughout the business day are uncoordinated chaos produced by the actions of millions of participants acting independently without any schedule between them.
But in that “chaos” you are describing, there are very identifiable, and exploitable repeating patterns.
The market can’t help itself.
Perhaps we are talking about two different things. According to what pattern did the hedging U.S. firm place its trade in the example I gave? Do banks or governments say: “Wait, we can’t place this order until 4:23 in order to stay on the pattern the global participants in this pair have agreed upon.”? If there is not a coordinated pattern recognized and adhered to by shoe shoppers in one small shop, how on earth could there be a coordinated pattern among the many thousands of participants from around the world in the EUR/USD market?
Let’s break this down a bit.
Firstly, no U.S. firm would EVER place an order based on time. They don’t care if it’s 2:37 AM on a Thursday morning, or 5:29 PM on a Sunday afternoon. They are interested in price. They need a certain number to open their hedge, or future, or forward trade. So for that reason, we have to throw time out the window.
Armed with that info, let’s stop looking at the micro of the single U.S. firm looking for a certain number, and look at the macro rest of the equation. That would be rest of the trading world. When that $100,000,000 trade hits the market, all hell breaks loose. At least temporarily. Price moves up, or down rapidly as orders get filled. There are levels hit along the way that other US, or UK, or Swiss, or German, or Aussie firms have orders at. Eventually price hits an order big enough that it can’t fully fill at once. Price bounces because there’s not enough traffic the other way to handle it, and stalls out.
That happens every day. And many many many times in a day. If price bounced at a spot once, it is the higher probability that it will do just that again. Those actions turn into repeating patterns on a chart. What seemingly looks like chaos is actually rather orderly. I don’t see the one minute or five minute charts as “noise”. They all have a story to tell, and sifting through the various time frames can certainly divulge a lot more information than they hide.
Exactly what I am saying. Concur.
We can say that when another firm places a later order that they are doing the same as the first firm in the same way that a second shoe buyer does the same thing as the first buyer. But, that is not a pattern scheduled on a clock such as the OP which said:“on the 1m, I noticed that peaks and valleys tend to form about every 3hrs”. The next move toward or away from some given level can occur in 3 seconds, 6.2 minutes, 3.78 hours, or 19.876 days. The time, the duration, and the magnitude of those changes in price and the events that produce them are completely uncoordinated and unscheduled. As you said: “They don’t care if it’s 2:37 AM on a Thursday morning, or 5:29 PM on a Sunday afternoon.”
While we know that a sudden move in a certain direction at a certain time on the five minute chart means somebody hit the market with an order, this information gives us no indication whatsoever about what will happen to the price of that pair at any given time in the future. Are you saying you can literally demonstrate a pattern wherein peaks and valleys cycle at x hours?
There are some good books on this subject that deserve mentioning. The first I already mentioned; I am slowly writing a book report about it on my tripped out blog: Fooled by Randomness - Nassim Taleb. Of course, all of Taleb’s books are must-reads. [I]Antifragile[/I] and [I]The Black Swan[/I] are so necessary.
The second I am only halfway through reading and I should do the same for it: [Calculated Risks: How to Know When Numbers Deceive You - Gerd Gigerenzer](http://www.amazon.com/Calculated-Risks-Know-Numbers-Deceive/dp/0743254236) This one shows just how badly numbers can mislead us. There is a pretty solid interview with the author [here](http://tunein.com/topic/?topicId=89140551). And I would suggest a listen to all of Michael Covel interviews by the way. Gerd Gigerenzer also has some great stuff on Youtube about heuristics and how they work. His contributions to the trading mind are gold.Great discussion! however, I think it got skewed from my original question. My question was not when, it was how long. For instance Obviously I don’t think that there will be a valley every day at say 3 pm sharp, but there is a lot of instances based from my personal observations that a peak and a valley usually terminates around 2-3 hrs on the 1m chart, again, from my personal observations. Has anyone else made similar observations? or is that still in the randomness area in which trader joe is?
I think they are just random coincidences. There may be more activity at certain times of the day but, as Acid says, the volume trades can be placed at any time of day