- We will use the difference in ATR and between the 3-day run up into UK EU Referendum (UER) and the Scottish UK Referendum ( ) in order to forecast what we expect price action to show on the 22nd, 23rd and 24th.
2014 3-DAY EVENT (17,18.19)
1. 1-Period ATR for the 17th 18th and 19th was 110pips, 163pips and 241pips - average of 171pips
2. 3-Day range was: 280pips - 1.6240 to 1.6520
3. On the day 12noon Implied ATM vol 17th-19th was: 8.8% 79th, 8.01% 52nd, 6.97% 22nd
4. On the day 10-period Historical Vol was: 10.4%, 10.4%, 11.1%
2014 3-DAY LEAD UP (14,15,16)
1. 1-Period ATR for the 12th 15th and 16th was 73pips, 53pips and 149pips - average of 91pips
2. 3-Day range was: 150pips - 1.6150 to 1.6300
3. On the day 12noon Implied ATM vol 12, 15, 16th was: 8.82% 76th, 9.34% 87th and 8.45% 65th
4. On the day (12,15,16) 10-period Historical Vol was: 10.9%, 10.8%, 10.4%
2016 UER 3-DAY LEAD UP (17, 20, 21)
1. 1-Period ATR for the 17th, 20th and 21st was 195pips, 371pips 155pips - Average of 255pips
2. 3-day Range was: 580pips - 1.4195 to 1.4775
3. On the day 12noon Implied ATM vol was: 23.2% 100th, 24.3% 100th and 20.16% 99th
4. On the day 10-period Historical Vol was: 14.1%, 19.4%, 19.2%;
*2016 UER 3-DAY EVENT (22, 23, 24) FORECAST*
1. 1-Period ATR for the 22nd, 23rd and 24th FORECAST: `293pips, 1141pips, 250pips; (171pips/91pips)*255pips = average 480pips (average adj 340pips), SD of 500pips
2. 3-day Range FORECAST: +/-1100pips - 1.4600 to 1.3500-1.5700
3. On the day Implied/ Realised ATM vol FORECAST: Event has been implying anywhere from 30%-60% over the brexit 3 day period, with ATM currently trading at 26% already.
1. The price action forecast around the event suggests that we could see a 1100pip range over the next 3 days (22, 23, 24) - given that we dont know the direction of the range, we can assume a distribution of 1100pip +/- at the current trading price thus forecasting GBPUSD to trade anywhere between 1.35-1.46-1.57.
- Further, the model expects an average daily range of 480pips, with the vote day skewing the average significantly (1141pips), therefore i think a 340pip (average adjusted) daily range is more likely.
2. Combining the estimated distribution range of 1.35-1.46-1.57 with the standard deviation of the foretasted daily ranges = 500pips, the model ends up showing significant statistical relevance by backward validating itself e.g. +/- 2SD of the mean at 1.4600 is 1.5600 and 1.3600 (+/- 2*500pip).
Before knowing this the model had already forecasted a 1.35-1.57 range thus this is somewhat reassuring as the model held true when back tested using +/- 2SD. 2SD is significant as it accounts for 95% of outcomes.
- The model also estimates that the tail risk of a BREXIT would cause GBPUSD to fall -3SD which is down to <1.31 (1.46 minus 1500pips) - this is also somewhat close to what I would have expected the day after the vote.
*See the 22nd, 23rd, 24th Trading strategy post where I link this information to execution*