Saturday, July 04, 2009

What Are Reasonable Performance Expectations For Traders?

How do I know if I'm doing well--or poorly--as a trader? That is the topic of an insightful question from a reader:
The question that no one has answered to my satisfaction goes to the
heart of managing my expectations. It is very hard to know how well
you're doing, or what you're doing wrong, when you don't have a
standard by which to judge your performance. For instance, a trader
once told me that a good day trader should make money every week,
not every day, but every week. Somehow I doubt that such level of
consistency can be reached. But this has caused massive confusion
on my part. Without knowing what elite traders go through, it's hard
to judge whether I am doing something wrong, or whether I merely
lack patience to stick out through inevitable drawdowns.

So my question is...where can I find information of elite day traders,
in terms of their performance? Mentors are notorious for not
opening their books, so the only people I talk to are aspiring traders
such as myself. The answer to this question will at least give me a
standard by which to judge my performance. For instance if I know
that it's normal for a top daytrader to have a losing run of 'say'
two-three weeks, then I won't 'panic' as much if I've lost money
three straight days.

I address issues of performance--and reasonable expectations for a trader's learning curve--in the Enhancing Trader Performance book. That book is useful in that it tracks the phases of the learning process and what one needs to do to move from being a novice to being competent to becoming expert. Chapter Eight of The Daily Trading Coach also suggests specific metrics for evaluating performance and guiding self-coaching. This is important, because it addresses, not only whether you're trading well, but also whether you're getting better over time.

At the firms where I work, trader P/L is audited daily and is a matter of record. I can see verified statements of how traders perform. This is useful because, as the writer points out, false portrayals abound. Many are faking it, hoping that they'll eventually make it. One guru who offered his services made the claim that he made over a million dollars from trading every year for the past ten years. That sounded pretty impressive--until I learned that he was only toting up the wins from his day trades. Yes, he made more than a million dollars' worth of winning trades. What he didn't say is that he also made more than a million dollars' worth of losing trades! Deception abounds--

Of the traders I've known personally through my work, only one rarely has a losing week. He is quite disciplined in his trading and limits his trades, both in size and setups. That prevents him from losing too much on any one trade, and it keeps odds in his favor. Of the hundreds of other traders I've known well, every one has had losing weeks and strings of losing weeks. Indeed, it's not uncommon for excellent traders to have losing months or a period of losing months. Markets change, slumps occur. If you are profitable on 70% of your weeks, 2.7% of the time--or about once a year--you'll have a string of three losing weeks simply by chance.

It's great to make winning weeks your *goal*. And it's good practice to never lose so much money in one day that you can't have a green week. But there are plenty of elite traders who make good money and undergo losing periods. I work with them every week.

So what *is* a reasonable expectation? Calculate your returns on an unleveraged basis. If you're making twice the riskless rate of return after commission costs, you're doing very, very well. If the riskless rate of return is 3.5% on a 10-year Treasury note, then a 7% return, unlevered after costs, means that you (or a trading firm) could leverage that trading (assuming scalability of trading strategy) into attractive double digit returns that would reward the firm (and firm's investors) as well as the trader.

By calculating your return after costs and on an unleveraged basis, you focus yourself on realistic, risk-adjusted returns. Retail commissions make achieving such returns challenging. If I have a $100,000 portfolio and trade a 2-lot of ES once a day at $5.00/round turn, I'll rack up $2500 of commission costs over the year. That's 2.5% that I have to make simply to break even on the year with infrequent day trading! For a frequent trader, the commissions are so high at the retail level that sustained success becomes prohibitively difficult.

If you have a genuine edge, you should make a respectable return without leverage. That can turn into a superior return simply through the judicious use of leverage. But if you can't sustain an edge without leverage, leverage will magnify your trading flaws and invite risk of ruin. In my view, you're a competent trader if you can consistently cover your costs without resorting to leverage. You're a superior trader if you can sustain more than the riskless rate of return after costs without leveraging your capital.

For more on this topic, check out my post on trading metrics (and its links) and this one (and its links) on keeping score in your trading.
.

14 comments:

Bryan said...

Thanks Brett, that's really helpful. What you are saying seems applicable to longer time frame players too, especially the x2 long bond target.

Mayare Real Estate said...

Thanks Brett,

It is one the most important posts

Thanks

Noam

Fredy G. Presutto said...

Hi Brett,
When you say twice the riskless rate, you mean yearly/monthly/weekly?

Tom said...

Excellent post...but...
What I'd like to see is a performance curve for experienced traders. Something like: "The mean unleveraged performance, before costs, for experienced traders (3 years or more as a full time trader) is X percent, with a Std. Dev. of Y percent." There are ways to increase nominal returns (leverage) and reduce costs (lease a seat). Further, with training and experience, one can probably improve performance. But to stay motivated, one needs to know where they are on the curve. If I'm already in the top 10%, then I probably don't need to tweak my approach. But if I'm in the bottom 10%, then I need to change something.

Trader Kevin said...

"One guru who offered his services made the claim that he made over a million dollars from trading every year for the past ten years. That sounded pretty impressive--until I learned that he was only toting up the wins from his day trades. Yes, he made more than a million dollars' worth of winning trades. What he didn't say is that he also made more than a million dollars' worth of losing trades! Deception abounds."

That's hilarious. Reminds me of Unemployment Friday at the Board of Trade in December 1997. A floor broker in the 10-year T-note futures pit got a 5,000-lot sell stop from a large trading house located at 85 Broad Street. The clerk handed it directly to the broker and said, "You're limited on three thousand of these."

He nodded. "OK, got it."

Number comes out bad, stop is elected, sure enough the broker launches all 5,000, even though it was way below the limit price on the last 3,000. So he's short 3K in the hole.

Broker realizes his error and buys 1,800 bond futures to hedge most of his market exposure. He lost about ten Note ticks per contract unwinding his disastrous spread.

I went to a Christmas party that night and the broker was there. I heard him telling somebody about his day...

"Yeah, I lost a million-and-a-half bucks on an error, but made back about 600 grand trading bonds."

I laughed to myself at this revisionist history. He could have said, "I f'd up and lost about nine hundred thousand on an error," but instead he counted the gains on the bond hedge of his error as "trading profits"! I guess it made for a better story.

Many of the clerks at the party had received their year-end bonuses that day. Some of them got drunk and headed for the gambling boats. When I came in Monday I heard a couple of them had blown their entire bonuses.

Amazing that people who would spend all year fighting too-and-nail for small edges for their customers would then go to a casino and piss away their money giving up the edge.

OKL said...

Brilliant stuff once again!

nan said...

Dr. Brett. Great Post. For me this is a very hard issue. The firm I work with, gives me the opportunity to trade a lot bigger and have a bigger stop out than I actually need. This seems like a very good opportunity. However, if I apply those new parameters, the daily swings of my pl will grow exponentially %wise compared to the size of my account. Some people might do great jumping from 15-20 lots to 80-100 lots. I know doing things gradual is the best. But Do you have any kind of advice or therapy for making those big jumps in size more manageable?? I really want to take advantage of this opportunity but I don't want to wipe out my account in a bad month. Thanks a lot.

gggg said...

Having the "riskless" return as your standard isn't always appropriate. If I deposit $10K in prop firm, and assume I'm doing 1000 shares lot and have a net p&l of (say) $50 a day. So in a month of say 20 trading days, that's $1000. Well that's 10% a month - I must be a genius. That of course isn't the case at all! I manage my commissions and pay $30 on 100K shares and for me, if I've done 80$ gross that's 80 / (100K /2)=0.0016 dollars. That is - If I just traded one lot of 50K shares, and did $0.0016 on it, Id have made 80 dollars. That is far from good. As people make 0.01 on their total share size and do like 5 million shares a day...another trader I know does 0.005-0.007 and does up to 20 million shares a day. So in the prop world, your returns percentage-wise aren't really useful. What *is* useful, is measuring how many "cents" you have done comparing to your share size, how well you manage your commissions, what is your largest loss, your max daily loss, your win loss ratio etc etc etc.

And, returns are nearly always not scalable. If you make 0.01 trading 1000 shares lots doing 100K shares a day, that usually doesn't mean if you were trading 100K lots and did 10 million shares you would still make same 0.01. In fact, traders having the most defined entries/exits most of the time have difficulties crossing certain lot size barriers - a lot is due to psychological reasons, but a lot is also due to the fact that as your lot size increases so does the difficulty of risk management and overall it could just mean many things you are used to having work for you at smaller sizes, just wont work at all now....such as liquidity issues, commission management, and market impact considerations.

In the prop world - I have found accuracy, stability, and efficient risk management to be a much useful benchmark than your returns percentage-wise.

OKL said...

It is most useful to review performance metrics as it allows the 'internal observer' to ask more pointed and precise questions when the 'player' is in control.

I somewhat agree with gggg, that adopting a comparison on trading with a supposedly riskless rate of return is like comparing apples to oranges, it doesn't quite make sense.

Drawing once again from the analogy between sports and trading; let's assume that the trader is a basketball player that made $100k/yr while the supposed riskless rate of return is the benchmark for all workers in the country is $80k/yr.

Going by the above example, can we say that the basketball player did well because he's better than the average worker? Sure, its possible... but it won't make a damn difference to the basketball player because that's the wrong comparison.

Instead, the trader has to compare himself with other traders; i might even add, of similar make. For example, if the basketball player in question is a PG, it doesnt make sense to compare his statistics with overall player benchmarks or PF benchmarks; only with overall PG benchmarks does the comparison/target make sense.

If you ask me, the hardest part is finding similar styled traders and compare their performances... however, the classification itself can be a real headache as there are so many different markets, instruments, position sizes etc.

To further elaborate, I don't think anyone enters the trading arena with a mindset of "beating a riskless rate of return"; it won't mean a damn thing, not to me at least... some of the very passionate traders might even call it an insult, much like professional sports players regard performance comparisons with the average worker with great disdain.

Going back to what gggg said; that accuracy, stability and risk management is the key to higher performance in the trading sport... that I agree and i might even add that 'situation awareness' and 'correct execution abilities' are crucial in playing this trading sport.

If for proprietary reasons there's no one to compare to, then compete against yourself; create your own list of "Best Winning Streaks", "Worst Losing Streaks", "Highest Scoring Games"... ultimately, as Napoleon said, "The greatest battle is fought within,"

siew said...

Hi dr,

Can you elaborate on the unleveraged return idea, please.

I strated off with a $50,000 account in 2005 and is currently trading with a $200,000 capital trading index futures ie es, nk, etc. I am never in the situation of having to top up my account on account of overtrading and generate profit after all trading costs of $40,000, does that mean that I am generating a return of 20 %? Please advise . Thank you kindly

Curtis said...

I would like to see this concept elaborated, as well. There is another way to look at this question: if you a trade a system then your goal is achieve results within the range of what the system predicts. A lot of people come into market with ideas that sound good such as I'll risk only a couple points, cut my losses fast, only take 2 trades per day, never take more then 2 losses, etc, etc. But, if your testing and system do not show that these ideas either reduce risk or increase returns then it doesn't make any sense. So, this question may be an indication of a lack of tested system. The next question is then what metrics should a top notch system have. Good question.

Curtis said...

Another way to look at it, is say the average salary a trader could make would be 50k thus to overcome the opportunity cost should be able to make at least that much. Taking Dr. Steenbarger's minimum 100k starting capital then you'd need to net at least 50% to overcome the opportunity cost but given that you will get the risk free rate without doing anything then you need to make that and a premium/bonus for the risk that you are taking. So, I would say around 60k. Thus you'd need to make NET 60% from 100k to make everything come out. Given that most traders are not able to sustain 20% then this indicates that 100k may be a bit low when opportunity cost are factored in. It seems that around 500 to 700k would be a more realistic figure.

The equation changes quite a bit if one doesn't need to trade full time or if one can automate their system.

Michele said...

Yet another excellent, thought-provoking post. Thank you.

I am currently wondering about the emphasis on trading costs. Obviously, the lower you can make your costs, the better. But as long as a trade is profitable, is the cost of executing it that important?

We learn from watching Discovery's show about the late Billy Mays that they consider a product a success if they can make just 3 to 4 times their advertising expenses. If they have to spend $1M to bring in $3M, they're happy. That's a 50% commission!

Taking smaller profits on more frequent trades would seem to be less risky too, since it decreases the market time exposure of one's capital.

AgeKay said...

Brett,

I think you're wrong on a couple of points.

First with regards to retail commisson. $5 per RT is less than half a tick in ES. So even if you make only 1 tick per trade, you'll be fine.

Also if you do intraday trading, it is not unrealistic to make a few percentage points every day by doing a couple of trades risking 1% on each trade.