Revision?

Nick77

Private
Messages
14
Could someone please explain to me what the revision column is on the economic calendar? Are the "actual" numbers wrong and the "revisions" the correct figures? Thanks.
 
Sometimes, more data comes out that proves the original numbers were wrong. When this happens, the old numbers get revised at the same time the new numbers get released.

For news traders, this can be a problem if the revision is up and the new numbers are below estimates (or vice versa). The markets can have some very ... interesting reactions when there is a conflict like this.
 
Still don't get it. (sorry)

So is the number in the revision column the newer most correct number, or is it a deviation? Is the "actual" number changed when corrected information is released?

When there is a number in the revision column how do I use it?
 
It is a little bit confusing.

Just so we are all looking at the same things, I'll use a real example from earlier this week. Look at the New Home Sales numbers for January 28th. These are the numbers:
Actual: 604
Forecast: 645
Previous: 647
Revision: 634

Before the news came out on the 28th, there would have only been numbers for Forecast and Previous.

In general, if Sir Pips says that a deviation of a certain amount should cause a currency pair to move a certain number of pips, he's talking about the Actual number. What can happen to mess this up is that the Actual number hits his deviation target, but the Previous number gets revised (this happens at the same time the Actual number is released) in the other direction. When that happens, the market action can be unpredictable. On the other hand, if the Actual number deviates by at least the trigger amount and the Previous number gets revised in the same direction as the Actual Number's deviation, then the market may move farther than anticipated.

In our real world example above, the Revised number and the Actual Number were both in the same direction (down).

So, for those of you trying to trade the news, hope for either no revision or a revision in the same direction as the Actual number.

I hope this makes a little more sense now.
 
let me get this straight

so yesterday (Jan 29th) consumer confidence got revised.
Last time it was reported at 88.6 but yesterday when the new report came out they changed the previous number to 90.6 suggesting it was higher than they reported the last time. The new number was 87.9 suggesting it had dropped not 0.7 but 2.7 indicating things are worse than they thought.
Is that right?
 
Let me check... oooooohhhhh, you picked a really weird one.

Yesterday, before the report was released, the previous number was 88.6 and the expected number was 87.0. Then at 10:00 am (Eastern US), the new number came out as 87.9 (+0.9 deviation above expectations), and the old number was revised to 90.6 (+2.0 deviation above the previous report).

That would make this a VERY difficult trade. On the one hand, both the announcement and the old number had positive deviations. On the other hand, the big increase in the prior number means that the new number is actually a bigger that expected reduction from the previous month instead of an increase. The expected number was 1.6 below the previous number, yet the report that came out was 2.7 below the revision of the previous number (a -1.1 deviation from expectations). [I hope I got all the math right]

I guess this is the perfect example of "I have some good news and some bad news."

Thus, there are 2 ways to look at this. The first would be: "Yaaay! Both numbers in the same direction and bullish for the USD! The second would be "Oh my. The total change was more negative that expected. Dump those dollars and buy something else." The reaction to this kind of release depends on how the BIG players decide to interpret the results. If they disagree, then look for some very unpredictable market action.

Luckily, this kind of "double positive possibly being negative" doesn't happen too frequently. Of course, a positive deviation on the new number with a negative revision on the old number can happen, and can also lead to unpredictable reactions.

If you really want to trade news, you should hope for no revision, or a revision in the same direction that isn't big enough to make the new deviation look like a reversal.

Happily, not all numbers are subject to revision. Today's anticipated interest rate cut falls into that category.

I hope all of this helps at least a little. If not, maybe we could invite Sir Pips to give us a better interpretation.
 
Thanks

Thanks. I get it now. Have you found a good technical system yet? I read some of your other posts about the "curse of the Pharoah". I am currently looking for a good technical system. At least until I finish my Macroeconomics course this semester, than maybe I will have a better understanding to news trade. If you have or do find a good technical system please let me know.
 
If anybody wants to learn about revisions look no further than what happened with non farm payroll news yesterday. Dollar negative headline number but dollar positive revision number led to initial spike in one direction before it went the opposite way on revision number.
 
Back
Top