Intuitively, there would be more 1% move at 2000 than at 3000. My two cents.

It all depends if you believe in a normal distribution curve or not....

Sounds logical. 3000 wld the new normal. The profitability of the 500 compnies wld be at a new higher level. If the companies profitability wld be as solid /or as non solid as today, PE same level, 1% movement wld occur just like today. My five cents.

In my experience with DAX and FTSE MIB futures, algos usually work better if those parameters that apply to price differences (for instance: a threshold on the X-day avg ATR) are expressed in points rather than in percentance of contract level / value. This may suggest that a 1% move at 2000 is more probable than a 1% move at 3000. I believe this may be due to the fixed tick, that is the fixed dollar value per point as suggested by Scott. If the volumes intended as the number of contracts exchanged do not depend on the contract value, then it may makes sense to to calculate the price movement probabilities in ticks. But as ticks have a fixed value, then these probabilities expressed in % are skewed in favour of the higher values as the contract value decreases.

A quick glance at Keltner bands on my ES monthly chart, which are based on range, is showing very little difference between the ranges at the 2000 and 1000 levels, going back 18 years. I was surprised, having expected to see range scaling.

Your assumption makes sense as percentage price changes are stable over time (independent of the price level) and not price changes. See the following link for the sp500 since the 50ties, upper chart daily close to close price changes, bottom chart daily log price close to close changes (for the use here log price changes are identical to percentage price changes):

http://www.ustermetrics.com/downloads/diff_vs_return_sp500.png?attredirects=0

incorrect

It's a valid assumption. It's not right nor wrong.

Check out Hauser and Lauterbach (2002)

It should not make a difference. The futures market follows the cash market +/- the basis , otherwise arbitrage opportunities arise. So at 3,000, a 1% move in the cash market should have the same probability as 1% when it was 2,000. Futures should therefore follow this law.

It could be 30 point if we assume that all traders became 30% more rich since 2000 and shrink their limits. But I think it's not. Thus risk in money should be the same. IMHO But on the other hand a volatility is driven by emotions, not money.

Why do I see no talk about implied vol !?

No

Short answer: it depends on your (trading) time frame. Does your trading model look to make ticks, or does it try to make a percentage on capital? Intraday futures traders care about ticks, and that number doesn't change. Equity intraday traders can fall in this camp as well, as they often find similarly priced names.

No, it looks good on paper, but this doesn't work on paper. Too many variables that change the lineage.

Hi guys - thanks for the responses on this topic - it's without a doubt one of the more divisive topics I've discussed with traders and investors . Most experienced traders seem to believe that the fixed point value of futures contracts changes this logical assumption, while the investor types think not.

It's 10%. However the ATR is 33% smaller. It's relative to It's denominator. What's your point, as it is very interesting

Typo 1%

Mostly a conceptual question that my team has kicked around internally and I was curious as to what others thought. Not considering the correlation - even if it is a mild one - can impact trade frequency, returns, risk management etc. when designing strategies.