Jump to content

ATR in Day Trading


Recommended Posts

Here some of my thoughts on the importance of ATR (Average True Range) in Day Trading.

Here's my thought process:

  • In order to make profit you need volatility
  • Volatility is the change in price of an asset over time
  • As day traders we're interested in volatility per day
  • To measure this I take the true range over one day, which is effectively the high of day minus the low of day and take the average of that over 200 periods
  • If we assume we only take one position on a given day, in theory the max profit I get is when I buy exactly at low of day and sell exactly at high of day (or short-sell at high of day and cover at low of day)
  • I need to subtract the spread from this max profit
  • The margin factor requirements from IG define how much I actually can make in terms of £. Higher margin factor means lower qty to trade with, means lower profit, even if Average True Range is high

 

I pulled some data today to find assets which fulfil requirements below

  • low spread
  • low margin requirements
  • high average true range

I looked into Indices, Crypto, Metals, Currencies, Commodities and Shares. I didn't fully automate the queries, so just pulled some snapshot data in the below. There might be some nuggets I missed, if you spot any, please let me know.
(All profit calculations are based on a hypothetical £10k account - for larger accounts things may look differently because the margin factor rises in tiers for large accounts)

Indices

IG has a few indices at 5% margin factor, which outperform the rest. Only exception is Australia 200 which is hanging behind the China 300, while the latter has 10% margin factor. Most profitable ones to trade would be NASDAQ, NIKKEI and DAX.
Interesting finding for me as currently trading the FTSE 100. Might give NASDAQ a try.

image.thumb.png.65017305db34416e70a789bc6b5b52d1.png

 

Crypto

Perform worse than Indices, because of 50% margin factor. ATR is much higher compared to Indices though, so if IG lowers the margin factor at any time in the future, these may become interesting.

image.thumb.png.1f24e4fcdfb7aa0e3aadb8026c3f56ec.png

 

Metals

Nickel and Spot Gold seem to do well. Gold because it only has 5% margin factor. Nickel because it has a large ATR of 2.8%.

image.thumb.png.f744a7ef318641d9dc673eaca3bd7a47.png

 

Currencies

There are only few currency pairs which have a margin factor of 3.33% and a few more with 5%. Those perform better than the rest with 10%. ATR is relatively low here. Best ones I could find are GBP pairs like GBP/JPY, GBP/ZAR and GBP/CHF. Might be related to Brexit and high volatility in GBP at the moment?

image.thumb.png.872c27a10de314fa61cabade68d1c224.png

 

Commodities

Surprisingly perform relatively well. Carbon Emissions, Natural Gas and US Crude at the top spots. 10% margin factor with relatively high ATR of around 3%
I might give those a try.

image.thumb.png.5f90705c8751acc2034eb26f51aec5ee.png

 

Shares

Now shares are a little bit different to the rest, because they can be very volatile at times and don't move at all at other times.
The best bet might be to find shares which were recently falling sharply aka trading well below 200 EMA.
IG then increases the margin factor, but that might take some time. So if you find a stock which recently fell sharply and you can get in before IG increases margin factor you have insane profit % of 50%+ like Metro Bank and Kier Group in the below.
Difficulty here is that the True Range has huge swings itself and you need to time it right to get on a big move.
That's why I'm sticking with Indices at the moment.

image.thumb.png.b51e6fd853c7db2954c57487c68f734d.png

 

So, if we ignore shares for a moment, the top 3 assets to day trade according to this theory would be:

1) Carbon Emissions

2) US Tech 100

3) Natural Gas

 

What do you think about this approach? Does it make any sense?
Commodities seem to be doing well in this approach - has anyone in here experience trading them? 
(US Crude and Iron Ore would be place 4 and 5 - that makes 4 out of the overall top 5 being commodities and 1 Index)

Link to comment

I like your work.

I am a retard so I don't understand 'volatility' in the same way.  To me volatility means chaotic movements in both directions, which is less desirable than a sustained move in one direction (the direction that you have bet on).  This is why more people have success with betting on an individual share price than day trading/minute trading on the 'great casinos'.

Link to comment
10 minutes ago, dmedin said:

I like your work.

I am a retard so I don't understand 'volatility' in the same way.  To me volatility means chaotic movements in both directions, which is less desirable than a sustained move in one direction (the direction that you have bet on).  This is why more people have success with betting on an individual share price than day trading/minute trading on the 'great casinos'.

Volatility is a reflection of movement, it could be up, or down, or up and down.

Link to comment

Cheers for the comments guys.

In regards to @andysinclair question: You can pull a lot of data via the IG API, check API documentation here: https://labs.ig.com/rest-trading-api-reference
You need to be prolific with VBA coding though.
In this example I did the analysis half-manually - and only used a quick screener I set-up in PRT (ProRealTime) to pull the assets with the highest ATR compared to their price, to get a quick snapshot what we're up against here - then pulled spread, margin factor and true range from PRT charts/IG online platform.

 

As the methodology in theory seems to make sense, next stage is to drill deeper.
I'm planning to take the top 10 assets based on the above data-set and break them down by day of week and hour of day.

Rationale behind this is to 

a) know the best asset to trade at a given time of day. Assumption is first hour after market open of the respective asset, but would like to confirm that. (e.g. FTSE 100 8-9am; NASDAQ 14.30-15.30; ...)

b) being able to derive a daily target price based on the predicted true range on a given day. E.g today is Friday. Usually on Friday's the true range is smaller, so target would be smaller as well. Goal is to quantify this.

 

I'll update this thread once I have more insights.

Link to comment

Ok everyone, pulled some data. Will split up in several posts.

First of all, here is the selection of tickers I examined in more detail.

I scaled down the hypothetical account size to £300, as this reflects a more real life scenario than the £10k for me right now. Also for beginners £250 is the minimum you need to top up your SB account, so £300 is kinda where you have to start.

Few of the top tickers fall out of the list now, because you need more than £300 to trade them.

image.thumb.png.2ba849b871178cd7d953b1cb6b2e86be.png

So, I will be focusing in the following analysis on:

NASDAQ, Natural Gas, US Crude, Spot Gold, DAX, FTSE, GBP/US

Link to comment

First I looked at breaking down ATR by day of week.

Example for NASDAQ below.

Takeaways:

  • Monday seems to be the most volatile day with ATR max spiking above 400
  • However looking at the average, it appears to be quite flat over all 5 days, ranging around 120
  • The Median might be the best KPI to look at as removes extreme values from the calculation. If looking at the Median it turns out Monday is actually the day with lowest volatility (Median ATR at 80), while Thursday is the day with highest (Median ATR at almost 120). (Even if the max values initially suggest the exact opposite)

image.png.eadc17a86ac19589eff874ced091b6ad.png

 

Charts for the other tickers below, without further explanation.

image.png.50583f7d22361fceeacc8294646fc229.png

image.png.d07809f64454e0845d38ef0d8f493528.png

image.png.4435e08b71fa8515483038c328be3da8.png

image.png.0ddf57e2b59d66e0e6743c1bf531f3c6.png

image.png.0489fcd0fd1653885f0f52d1bddd0326.png

image.png.3833663cb91a6dbbfbe0515a3f6aa26c.png

 

 

Now looking at overlaying the Medians of the ATRs of the different tickers giving the chart below.

I had to normalise the ATR, because when Gold moves 20 points it's not equal to NASDAQ moving 20 points, because of how big you can size in. Trading Gold, for example, can be sized 6.1x as high as NASDAQ, hence the ATR is multiplied by 6.1 as well (this is because margin factor and price are different for each ticker).

Best ticker to trade per day:

  • Monday: US Crude
  • Tuesday: US Crude or DAX
  • Wednesday: US Crude or Dax
  • Thursday: NASDAQ
  • Friday: DAX

 

image.png

Link to comment

Now to the interesting stuff. Breaking ATR further down by time of day.

NASDAQ below.

Key Takeaway:

  • Trade best between the hours 14 and 15 (first 90mins after market open 14.30-16.00 as expected)
  • Again Median should give a better indicator than Average as it removes extreme values, so we're gonna focus on that for the comparison

image.png.a3c584f1aefdd8e6b089419aa864985a.png

 

Same chart for the other tickers:

image.png.8ab3abdec47920a25549c600010d9fc8.png

image.png.4c6ab0f8474d891c51b311eea4076ca3.png

image.png.a415f9aa0067b3af74080b78675f44ce.png

image.png.ec07cd35c94e477cdf837c6066afc202.png

image.png.ee7f4d90a24284d86b81645b081b4726.png

image.png.e13472ff9b95009b10f2a96b0c3c7c0e.png

Key Takeaways:

  • Indices are best trades in the first 90min after market open or in the last hour before market closure
  • Gold can be traded throughout the 24h-day with spikes at market open Asia, Europe and US
  • GBP/USD ATR remains high during entire EU market hour session (8.00-16.00) with peak at hour 15
  • Natural Gas and Crude Oil seem to follow US market hours, but have a very steep curve. They only really seem to move between 14.00-16.00

 

Now to the comparison, again with normalised ATR values.

  • Trade Gold at night (if you fancy getting up at 2am)
  • Trade DAX or FTSE between 8-9.30am
  • Trade NG, Crude or NASDAQ at 2.30pm-4pm
  • Trade Crude at 7pm-8pm and NASDAQ at 8pm-9pm (if you still haven't got enough)

image.png.21e4f448818c2d825dff8b85dd8f5125.png

 

Link to comment

For me, this means, I will keep trading FTSE in the morning at 8-9.30am

Additionally, I will start trading NASDAQ at 2.30pm-4pm.
Occasionally, I will have a look at NG and Crude in the afternoon as well and see if somethings going on there (especially on Mondays where those seem to have the highest median ATR in this group)

 

Would love to hear your feedback!
Who thinks this is all utter rubbish, please let me know :D

Link to comment
29 minutes ago, DSchenk said:

For me, this means, I will keep trading FTSE in the morning at 8-9.30am

Additionally, I will start trading NASDAQ at 2.30pm-4pm.
Occasionally, I will have a look at NG and Crude in the afternoon as well and see if somethings going on there (especially on Mondays where those seem to have the highest median ATR in this group)

 

Would love to hear your feedback!
Who thinks this is all utter rubbish, please let me know :D

Absolutely Outstanding work there bud..............................WTF are you not already winning?? :D

Sorry, I've not had notifications coming up for this but i do now.  I've had a Viral Infection all last week and spent most of it in bed.  Let's see what this week bring with a small account.  Good Luck.

Link to comment

 

12 hours ago, DSchenk said:

Would love to hear your feedback!
Who thinks this is all utter rubbish, please let me know :D

Great work and thanks for sharing, not rubbish at all. As the data shows there is a lot of activity on market open (London 8am and US 2:30pm uk time). I tend to be a bit cautious on the open and for a short time after, all the overnight orders are processes sequentially on a first come (onto the order book) first served basis and then there are  new orders reacting to that so there can be a short period of price whipsawing. Also it's not unusual to see an attempt to check the pivot with a view to pick up more support before heading in whichever direction.

Link to comment

Thanks for the feedback.

I had a go at trading NASDAQ a bit and boy, I have to say, it is on a whole different level to FTSE. More one for Adrenalin junkies.
20 Points, 40 Points even 60 Points in a matter of 30-60 mins can be monitored frequently.

Not sure what the best risk management is on NASDAQ yet.

On FTSE I'm using:

  • risk per trade 5 points
  • profit target 5-10 points per trade
  • daily max risk: 20 points (4 red trades - I'm out)
  • daily profit target: 20 points

FTSE Median ATR in first hour of market open is at 34 points  and 2nd hour is 27 points.

On NASDAQ it looks more like:

  • risk per trade 10 points
  • profit target 10-20 points per trade
  • daily max risk: 30 points (3 red trades - I'm out)
  • daily profit target: 30 points

NASDAQ Median ATR in first hour of market open is at 43 points and 2nd hour is 42 points.

Doesn't look like much on paper, but you can feel the difference when trading.

Link to comment
12 hours ago, DSchenk said:

On FTSE I'm using:

  • risk per trade 5 points
  • profit target 5-10 points per trade
  • daily max risk: 20 points (4 red trades - I'm out)
  • daily profit target: 20 points

This is how I used to trade with a 2k account.  But I would risk maybe 3pts or £15 and take 3 pts.  I figured I'd need 10 trades with 70-30P/L.   In the end I realised with only 2k, I'd never be able to make the profits needed to sustain unless I traded for hours.

Link to comment

@nit2wynit That depends on how you define "sustain".

As you see from the analysis, on average the FTSE makes 34 points in the first hour.

If you are able to capture 20 points out of this with a £2k account, that's £100 profit per day /per 1 hour of trading. (qty traded: 5). That's 0.5% return on equity.

Now that's month 1. Month 2 you are at £200 per day. Month 3 at £400. (in theory, if you don't have any losses and make consistently the 20 points per day)

I have yet to find someone who is doing that kind of performance on stocks with gap-and-go setup or similar. Surely, once a month you may be able to hit a home-run with stocks and make 20%-100% return on equity on a single day, but what about all the other days? Are you able to not trade these days or trade and not lose? That's the tricky part

Link to comment

If we're talking about FTSE or Dow or G30 at Market Open then we're still talking about Breakouts, aren't we? 

There's been a few passers by on here saying they made most of their money on the FTSE Market Open for sure, but for me at least it still comes down to Risk/Reward/Account size/PPP/Time.  That same account size of £500 or 1k or 2k gives you more buying power in Shares.   Maybe it's Greed?  or maybe it's simply Time/Effort vs Risk/Profit.  It comes down to your goal in the end.  As  Photographer etc. I can make £100ph in the real world with zero risk, but on Shares I could potentially find a stock for £50 pp and hope it makes 5-10 points in 1 minute.   Side by Side, my max PPP on FTSE with 2k is £5.  This morning is a good example from 8am.   There's 100pt drop.  Cool that's £500 in 4 hrs or so, but would you rather make £500 in 4 hrs or £500 in 5 mins?
 

Link to comment

So far there's no strategy here...only some numbers and charts on a sheet :D

We found out when certain indices are most volatile and therefore best to trade. Question is now, how to capitalise on this volatility...

I don't have an answer to that one yet, but will make sure to launch a £1000 trading course once I find out :D

Link to comment

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...
us