How to be a Guru in 10 easy Steps

I'm starting to wonder if the phrase "positive expectancy" should become a new saying for Gurus. Sounds guru-like if you are talking to newbies and non-statisticians. Of course I have a positive expectancy on every trade. sweet. lol.

I don't think I added this yet; however, to all my guru-wannabes, try this one out for size.

Suggest 10 stocks, 5 short and 5 long. Play options on a few of them, out-of-the-money.

Then I am almost positive you will get ONE big winner. Then market that winner for weeks.

Another crappy saying is "high-odds" trade. They have no idea if it is high-odds or not. If so, they would show how they played such EASY trades in the past and made bank. They are simply guessing resistance or support holds. Sometimes it does.

It's funny, I'm sure all gurus think this thread is a huge waste of time by me.
HOWEVER, I am trying to help a few new traders understand how to avoid the used car salesman of our industry. They are peddling snake oil and must be avoided. It would be analogous (gurus love analogies) to going to a car lot but picking up the book "how to deal with a douchenoggle car salesperson" right before you entered the lot.

The irony is I am now the guru on 'guru-haters', so I guess I need to admit I don't know anything and all my guru guessing could be 100% wrong and I suck at my analysis most of the time and if wrong I will attack them again, etc...

LOL

 

Here is a TEST. How many guru-like statements can you name in the following article?

LOL

---------------------------------

BUILD YOUR DREAM!

Are you a successful active investor?
The active investor knows that when handled wisely, his or her portfolio gains can provide an extremely enjoyable lifestyle, along with a secure financial future!

The key word, here, is “active.” No longer content with the sometimes good, sometimes not-so-good returns, the successful active investor knows that “pot-shot,” buy-and-hold methods are old-fashioned. In fact, in this market, they can be downright dangerous to your wealth!
That’s why the successful active investor earns consistent profits by buying high-quality stocks and exchange traded funds that show the best potential of increasing in price value.

To maximize gains--and avoid the pain of profit-eating downturns-- the successful active investor holds winning positions for a time horizon of weeks to months. He or she manages risk sensibly, and steers clear of common mistakes made by most investors.

When Mother Market exhibits a volatile mood, the successful active investor calmly steps aside, pocketing gains until the market issues an “all clear” signal.

YOU CAN BECOME A SUCCESSFUL ACTIVE INVESTOR!

By using our simple, yet powerful approach to investing, you can be a successful active investor. That means you can enjoy steady gains, low risk, and feel confident and competent in your ability to control your financial future!

 

It is


RN

 

I would expect any person looking into a trading system or considering investing in any advisory service to understand at least a little probability theory. Positive expectancy without an explanation of what it is and expecting people to accept that you have positive expectancy with your system is the nature of gurudom, yes, but it is a great disservice for those of us that actually know what it means to dismiss it as guru-like offhand. Positive expectancy assumes that after completing your backtest you know your probability of winning times the percentage win less the probability of losing times the percentage loss should be greater than zero. To shorten this calculation, you could just say that a system has positive expectancy.

Indeed I've seen it used as proof of a system maybe incorrectly, but it takes third party verification to validate any trading system's claim, and is the main reason I can ride on anybody that comes onto ET and immediately expects recognition for free. Common objections when you start questioning a system include that they saw it work well in other places, and I've done it this way for 10,20,30 years despite affordable trading only being available since around 1995.

 

OK, I see your point. Let me ask a few questions:

1. How much data do you really need to make the results meaningful?

2. Is a sample bias likely?

3. Isn't there a random risk factor involved, so garbage in and garbage out?

4. Does position sizing come into play?

bwolinsky, I don't mean you personally, since if you have been doing this for decades you understand the importance of reliable data. I'm more curious about how the "gurus" use this term to simply sell systems; additionally, when a guru uses this i will know what questions to ask. Thanks.

Math seems pretty finite and the markets have infinite "tails" scenarios that could cause havoc. Moreover, a short-term edge isn't the same as long term, or is it?

 

I definitely recommend the phrase "money management" to gurus.

It's such a gurulike phrase. Every newbie will be in awe if you throw the phrase around freely.

"risk management" may have the same effect.

 

You need at least 30 trades of buy and sell, not buy buy buy buy sell sell buy buy sell, which is only 1 trade no matter how many legs there are in the trade. I tend to believe only an entry and exit constitutes a system. As to length of time I'd say at least 1 year. Meaningful results must be taken in context with primary considerations given to the Sharpe Ratio. Sharpe above 2 is always what you should start with, especially for shorter time frames under 2 years. Sharpes out beyond 2 years tend to be lower than that, but that's all right because we're talking about risk adjusted returns and sharpes above 2 are never sustainable. 1 might be, but try to find a mutual fund above 2 in the last 3 years, and I guarantee you won't find one at the 5 year mark.

Only if you have picked the market because it went up, or because it went down, or, the worst case being that you were too stupid to look at your data in the first place. There are times when you want to test your system against these market conditions, but generally if you picked your timescale specifically for the reason there were strong trends in one direction, you shouldn't expect the system to work going forward. If you recently happened on a program that showed good results in the last year, yes, you might want to wait a bit and see how it does, but the idea behind the sample bias has more to do with what your system is based on than it is about what the results were. As you said you can just say buy at stop dow 1000,2000,3000 but this is obviously in hindsight and you'd never want to trade it. I find in my system development depending on when you start is just as random as what you optimize on. The real bias comes when you start optimizing without out of sample data. I'm not one for random data, because random data does not behave like real market data.

I don't believe so. Random is what a system without positive expectancy is. Garbage in I find only happens with limit order scripts with bad data as well as on fundamental data systems. I know systems with multi-year track records are far more likely to be profitable in the future than one based on less than a year worth of data if they also lack a sufficient amount of trades, and generally that sufficient amount is at least a normal distribution of 30 trades.

Position sizing shouldn't be the basis of a system, and a prime example is any martingale system that never has any edge. Sizing is almost always 100% of equity for my systems, especially on ETF's and futures, you can either put all your money in an ETF or a synthetic 100% of equity on the futures contract. Position sizing for systems is about what the optimal risk reward tradeoff is. I have systems that have optimal risk reward ratios at 96% of equity, and some that only require 20%. Depending on your tolerance for risk is how position sizing comes into play. It's about your comfort level and tolerance for risk. As I said, it's never the basis of a system.

 

Thanks.

Have you ever used the Sortino Ratio?

http://www.investopedia.com/terms/s/sortinoratio.asp

For small accounts, Kurtosis I would guess is the real nemesis, or people that trade penny stocks.

Is that the same with systems?

I do see that on covestor there are traders that have sharpe ratios in the thousands.

LOL

 

I don't believe the sortino ratio is as useful at capturing volatility as the Sharpe is. While I can appreciate pretending upside volatility is not volatility, the statistician in me knows that's not the case.

Covestor in most cases has a limited time period for most of the members, so these go to unreasonably high sharpe ratios, especially over short periods of time.

The problem with small accounts isn't really that fat tails are the biggest headwind after commission, but, small accounts by themselves are undercapitalized and the extra commission paid is enough to make even a more robust system look worse than it is on a system without the extra costs. That and on a smaller scale you can't perfectly size your positions. The case in point is when a small trader tries to buy 108 shares of a 46 dollar stock with just a $5000 account and the extra .69 of shares left over if the system is based on 100% position sizing is actually trading at a smaller percentage than the system was originally designed for. The 100% of equity you were going for becomes 99.36% and this has the effect of decreasing potential long term returns to below expectation. Whereas if you had $500,000 this 10869 share order is closer to the 100% of equity position sizing you had planed for. In this case, rather than only 99.36% you have 99.94% of equity used as you had designed.

So, yes, I'd say that problem is worse for undercapitalized accounts for the reasons I described above, but definitely not as big of a problem for well capitalized traders. It's interesting you mention penny stocks, because I guess the position sizing problem that gives rise to the added volatility is mitigated with a small share price. Either way, if you're undercapitalized you're still going to have that added shortfall of performance that really does affect your expected distribution.

 

Today's Guru-ism:

Even though you never said "1018 support and 1030 resistance", put in your newsletter the following,

"The S&P hit obvious resistance at 1030, falling exactly to our support level of 1018."

LOL

it's such crap, but maybe you get one idiot that is just following and thinks you are a real guru.

Pivots are good too. Use daily, weekly, monthly and yearly.

We now have 20 levels.

I'm sure the S&P gets near one of those 20 and does something.

And if it just blows through one of those levels, make it seem like you knew that would happen.

"The ES traded right to our daily pivot level, first triggering stops then blowing through this level as predicted."

LOL