How much is an ATS worth?

Status
Not open for further replies.
Quote from dom993:

The same question goes for anything really ... whether football or backet-ball or hockey players, objects like boats / cars / bikes / pens / computers / phones, books & software, food, art pieces, and of course companies.

It really takes an open market to "objectively" put a price tag on something, with competition & customers in large enough quantity for prices to find an "objective" equilibrium.

It doesn't seem to me that ATS have an open market. Sure, there are numerous ATS providers, but are there many customers? Which ATS & providers are the market leaders?


When it come to a particular individual & a particular object, the "value" of the object has 3 components:

- usage value: the value the object can (help) generate when using it
- exchange value: the value a similar object can be bought / sold on the open market
- sentimental value: the value the owner attributes to the particular object for personal reasons.


In the absence of an open market to determine the exchange value, an ATS value should be based on its usage value, a formula taking into consideration (per unit of trading, ie. per contract / 100 share / ... ):

1. ATS estimated expectancy
2. ATS estimated max.DD over its intended lifespan (capital at risk)
3. ATS estimated P&L over its intended lifespan
4. ATS estimated probability (distribution) of "success" over its intended lifespan

1. / 2. / 3. can be projected from past results (backtesting or live, best being live results matched by backtesting results for the same time period), as long as those past results are trusted. Conservative & nominal estimates should be made.

4. is where the most risk lies for the user. A careful analysis of the ATS internals might help mitigate that risk, but no matter what, any trading system's edge (fully automated to fully discretionary) can be voided overnight - or longer periods time - by regulatory, exchange, political, economical or technology changes.
Excellent post. :cool:
 
Quote from kut2k2:

Excellent post. :cool:

Not so excellent. The pricing of an ATS suffers from adverse selection. In fact, it's so bad that with pedigree, an ATS is worth effectively zero.
 
Quote from dom993:

IMO the number of trades is not enough in itself ... there should also be a condition on the time period those trades cover. I find 10+ years a pretty good benchmark (where applicable), and I would now completely discard anything below 5 years.

For me, those figures apply for the backtesting part of the ATS. A one-year & 100+ trades live track-record, matching the backtest for that same period (95%+ of "identical" trades), would be enough for me to validate the entire backtest (assuming the performance characteristics for that one-year period are close enough to the prior part of the backtest).
This I completely disagree with. What matters is the number of trades aka your sample size. Time span is meaningless.
 
Quote from dom993:

Winrate is really meaningless by itself. The first "single metric" I would suggest to look at is the profit factor P/F.

But single metrics are quite deceptive, especially for large backtests, as they completely hide the variability of performance. Looking at the distribution of a particular metric on a year-by-year & month-by-month basis give a good idea of the performance variability.
Of course winrate is meaningless by itself. I'm not suggesting a pricetag based solely on winrate, I'm just trying to get a price factor based on winrate. You suggest PF but you don't suggest a way to evaluate it as a price factor. I don't know what all the "relevant" performance stats are and that should be part of this discussion, but the idea is to come up with a method of evaluation that includes all the relevant performance factors, which may or may not include winrate or PF.

The beauty of looking at winrate is that it allows pricing all the way from worthlessness (winrate = 0) to the Holy Grail (winrate = 1).
 
Quote from danielc1:

Van and Ken talks about a SQN number:

System Quality Number: (Expectancy/Standard Deviation R) X square root number of trades.

If you can agree with that ranking for how good a system is, you can go to the next step and determine how much it is worth.

IMO: No automated system is worth money up front.
Who are Van and Ken?

I like what I see of this SQN. Is there a derivation posted somewhere?
 
Quote from kut2k2:

Who are Van and Ken?

I like what I see of this SQN. Is there a derivation posted somewhere?

Van Tharp & Ken Long.

The SQN "formula" can be found in a few on Van's books.

The formula suffers from using the number of trades, IMO.
 
Quote from kut2k2:

This I completely disagree with. What matters is the number of trades aka your sample size. Time span is meaningless.

So you don't believe in regime shifts... interesting.

I'm guessing you are new to this?
 
Quote from dom993:

Van Tharp & Ken Long.

The SQN "formula" can be found in a few on Van's books.

The formula suffers from using the number of trades, IMO.
Thanks.

Here's a preliminary pricing factor. Feel free to dissect it. :)

(Winrate/(1 - winrate))*(PF - 1)*SQN/sqrt[1000]
 
Quote from CT10Gov:

So you don't believe in regime shifts... interesting.

I'm guessing you are new to this?
So you don't believe in different time frames ... interesting.

I'm guessing you are new to this.
 
If you were talking about valuing a large cap business for sale, then valuation would proceed along any of the following lines:

- Net Present Value of discounted future estimated net cash flow (after any and all costs, taxes, interest payments, etc where relevant).

- Comparable transactions (i.e. looking at publicly available information about comparable corporate transactions or asset deals … what was paid as a multiple to last year’s cash flow or income, and/or of the year before that, and/or of next year’s projected, and/or as a multiple of the number of strategies purchased, etc …)

- Comparable trading multiples (i.e. what are publicly traded comparable companies valued at based on their stockprices, again as multiples of the above criteria, etc …)

- If there are enough publically traded comparables, you may even be able to measure and estimate a cost of capital that you could use in Discounted Cash Flow calculation above …


… none of which is easy to do in this case.

Are there any publicly traded companies that are essentially a bundle of strategies and folks who know about and operate those strategies? If so, and you can find out what they made last year, then you have a multiple of value to last twelve months income or cash flow or whatever … and you can apply this to your own case (i.e. against what you made last 12 months)…

At this point however you have to recognize that you are not a publically traded large cap business …

- The buyer's ability to turn around and sell on your strategy to someone else is not guaranteed … big discount.

- You are talking about a single strategy (not a large bundle … so there is lack of diversification etc) … big discount.

- You are presumably planning on selling the strategy without the guy who knows and operates the strategy (i.e. you)… another big discount.

- Is the track record a real one or a backtest? … if the later, I suspect a buyer would generally apply the biggest discount of all … to something pretty close to zero, I suspect.

The best chance of a deal would be one where you address the above points (i.e. bundle the strategy with others, stay involved, and build a track record first)… so basically letting someone farm in to your strategy/strategies while you continue to trade it/them … much like the prop deals that another poster mentioned above ... or starting a hedge fund ...
 
Status
Not open for further replies.
Back
Top