Feb

18

More on the points vs % argument. % vol or Vix is misleading and inaccurate measurement of vol. A better measure is abs vol in points/$ because we live and measure ultimately in dollars.

Accordingly at 7000 abs vol is 350% of what is was a few years ago at 2000. Trading has to adjust accordingly to maintain the same portfolio volatility of returns. Thus leverage, targets, systems, time have to adjust to match.

Adam Grimes disagrees:

because we live and measure ultimately in dollars

This does not strike me as coherent. Returns are the only reasonable way to understand market movements. Just imagine a portfolio of two assets, one at $1 and the other at $100,000 (or other arbitrarily wide handles). The only way to think about those is to normalize for price via %, so it follows that volatility would be equally incoherent measured in points. (Sorry for the Finance 001 example, but I think 'Explain It Like I'm Five Years Old' cuts through a lot of confusion.)

% and vol measured as vol of %'s (i.e., returns) is the only thing that makes actual sense here unless I'm misunderstanding the argument. What am I missing?

Accordingly at 7000 abs vol is 350% of what is was a few years ago at 2000.

Scratching my head here. "So what?" and "of course" are the only things I could manage to say here.
I'm probably missing something, though. What is it? (btw VIX sucks as much as any other clunky measurement of implied vol. I'll agree with you on that one, but I don't think that's your point.)

Zubin Al Genubi responds:

1 ES contract used to move 7 points as an average range 20 years ago. You made or lost $350. 1 contract now moves 50 or 100 points a day, same percentage, but your account is up or down $5000. The volatility in your account in dollars is higher than 20 years ago. If you lost 1% in 2000, its $600, but 1% now is $3500 per contact. You don't see too many 2% days like before. Abs vol is up while Vix or % vol is down. Its apples and oranges.

The trading style, research should change. Straight percentages for expectations, returns, targets don't work like they used to, especially using historical data. Silver now has micro contract because $5000 a dollar too is high volatility in dollars with $20 or more ranges. I guess I'm suggesting using abs vol as a better measure of vol.

Appropriate here, Feynman in 6 Not so easy pieces, cites Wyle on symmetry, "Suppose we build a certain piece of apparatus, and then build another apparatus five times bigger in every part, will it work exactly the same way? The answer is, in this case, no!" My point is the market does not the same way as it did 25 years ago in large part because it is bigger. The fact that the laws of physics are not unchanged under a change of scale was discovered by Galileo. He realized that the strengths of materials were not in exactly the right proportion to their sizes. [Ibid]

Adam Grimes writes:

I'm sorry, but I still find these points trivially obvious. Of course nominal price swings are bigger because price levels are higher, so of course holding a single contract would result in larger dollar swings. Who's holding a single contract? Position size takes care of all of this.

And I don't think the physical analogs add anything beyond confusion. Physical properties scale differently. For instance volume scales as cube and surface area as square. This is why we could not have a science fiction 100 ft tall lobster in reality…because of material constraints. There's no market analog to this. A 1% move is a 1% move. There's no hidden non linearity there.

If your claim is that markets don't move the same they did 25 years ago I would challenge that claim. What's the evidence for this? Statistically there's always the issue of non-stationarity but it seems to me you're claiming there's something more meaningful here. What am I missing?

Asindu Drileba comments:

I think Zubin is simply trying to say that he had found measuring volume in dollar terms (absolute terms) more relevant than measuring it in percentage terms.

Richard Hamming has an interesting talk, You get what you measure

Here us a good summary:

You may think that the title means if you measure accurately you will get an accurate measurement, and if not then not; it refers to a much more subtle thing - the way you choose to measure things controls to a large extent what happens. I repeat the story Eddington told about the fisherman who went fishing with a net. They examined the size of the fish they caught and concluded there was a minimum size to the fish in the sea. The instrument you use clearly affects what you see.

I for example, completely stopped measuring market returns in percentage terms a few years ago. I now exclusively use log returns. Why did I stop using percentages? The problem with percentages is that they are not equal to each other (ignoring the negative sign). (a 50% move) + ( a -50% move ) does not give you 0 in dollar terms. But what you get is 0% in percentage terms. Percentages are not symmetrical. Does this mean they don't measure growth? They actually do. But they simply should not be compared. As the absolute values may mean something different.

- A 0% return percentage may (erroneously)
indicate that you have broken even.

- The same 0% percent return may also show that you are actually loosing money in absolute (dollar) terms

Adam Grimes writes:

Of course log returns are well known, and this is more finance 101. There are several qualities that make them more attractive for some analyses. (just dont mix percents and log returns!)

But that's not the same as measuring market movements in raw dollars (which is the only reasonable companion to volatility measured in absolute dollars (or points).

And as for measuring what you see, methodology (and perhaps even experimenter expectations) greatly affecting outcomes and conclusions, we're on the same page there. This is fascinating territory for discussion and I'd welcome it.

But his point about volatility only extends to someone trading a single contract in 2001 and also trading a single contract today. That is irrelevant.

If there's something at work here and legitimately some way the market "doesn't move the same way" it did decades ago… I'm all ears and very interested. Always looking to learn more about what I don't know or might be missing.

Zubin Al Genubi does some counting:

There were 31 days in 2006 and 67 days in the last year with percent moves >1%. This is due to Big Tech being 32% of ES with higher beta and the speed and intraday persistence of algorithmic trading. Lastly, it 'feels' different. A 120 point drop trades different than a 13 point drop in 2002.

Stated quantitatively, now nearly half of our trading days have abs vol hi-lo >1% while in 2006 only a quarter of days had abs vol >1%. That is a big difference and clearly explains why trading is different (better) now. Today is just 1 of the many such days.

Cagdas Tuna adds:

All futures contracts of index products have adjusted to gross notional value of underlying stocks. At the same time VIX contract specifications and notional value it can represent almost remain unchanged. Although calculation method of VIX is the same, the number of futures contracts hedgers need to cover notional values they trade in underlying assets are totally different.

Adam Grimes writes:

Stated quantitatively, now nearly half of our trading days have abs vol hi-lo >1% while in 2006 only a quarter of days had abs vol >1%.

I'm sorry but I have to be direct. I find this annoying. You have moved the goalposts and are now making a completely different argument. You began advocating for measuring volatility in absolute dollars and now you are using a percentage measure. You are literally using the metric you said was wrong to support your argument that the metric is wrong.

Furthermore, your data are bad. It's simply a volatile measure. Volatility is volatile. Here's a look at the FULL history of the ES futures [first chart] (back-adjusted so this may not be fully accurate, but I think the % adjustment fixes the back-adjustment distortion) counting the number of trading days in each calendar year that had abs(high-low) / close > 1%. It's simply an unstable measure and I don't know what is to be drawn from this.

If your argument is that there are more days with "surprise" distortions, this is maybe nominally true, but better measured with better tools. I use a tool that expresses each day in terms of the mean average absolute closing difference [second chart]. Taking an arbitrary bound of 3 for that, the argument that this year is on track to show more surprise shocks than usual is true (but also reflective of a generally lower baseline).

You've flipped to percentage measures and we're left with hand waving "it feels different now." That's a different claim and it's qualitative, perhaps has value, but not in supporting your original claim.

I'll reiterate my position: percentage measures are the only thing that make sense. Point values are arbitrary and can be easily handled via position sizing (within the limits of granularity of instruments vs account size.) I don't think this is revolutionary or controversial. This is truly finance 101.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search