This course covers two of the seven trading strategies that work in emerging markets. The seven include strategies based on momentum, momentum crashes, price reversal, persistence of earnings, quality of earnings, underlying business growth, behavioral biases and textual analysis of business reports about the company. In the first part of the course, you will learn how to read an academic paper. What parts to pay attention to and what parts to skim through will be discussed here. For every strategy, first you will be introduced to the original research and then how to implement the strategy. The first strategy, Piotroski F -score will be discussed in detail. You will be taught how to calculate the F - Score and how to use this score in a strategy. This is followed by the next strategy, Post earnings announcement drift (PEAD).
Piotroski F-Score : Implementation - b
Loading...
来自 Indian School of Business 的课程
Trading Algorithms
294 个评分
从本节课中
Module 3 - Trading Strategy 1 - F Score
After completing this module you will understand the Piotroski F Score Strategy and the economic intuition behind it. You also be able to implement the trading strategy.
与讲师见面
Prasanna TantriSenior Associate Director
Center for Analytical Finance
Okay, let's do some bolts and nuts part of it, implementing the strategy.
So far, so good.
We have the score.
We have categorized companies into nine, eight, seven, six, five, four,
three, two, one, zero.
And now, we've decided that we're going to start going long on nines and
short the zeros.
Or supposed nines are very few in number, I already told you that it's unlikely that
you'll find large number of stocks, which are valued low by the market.
So if they are few in number, we'll do say nine, eight, and zero, one.
By the way, what's wrong with this trading few number of stocks?
Suppose you have only say four stocks with nine.
What's wrong with that?
Nothing is wrong, but that exposes you to idiosyncratic risk.
Something wrong going on with the stocks.
Please remember, these are Piotroski scores is not a gospel of final truth.
This is just a signal about likelihood of the company.
This is not the certificate that companies want to do well.
So it's quite possible that a particular firm does well in online parameters, and
some general and unforeseen hits.
Like we discussed in the demo.
It is quite possible for, say, an oil company,
which did very well up to, say, 2013.
When oil was touching 100 and crossed up all around 90s and hundreds.
Suddenly, the price falls 50%.
The [INAUDIBLE] may be good.
But that does not mean that the future performance also is likely to be good.
So if oil falls from 150 Petrosky or no Petrosky,
there is no way the firm is going to do well.
So suppose you're out of the four stocks that get a score of 9.
If three are from oil, then your portfolio returns will get severely compromised.
So you have to ensure that you have reasonable number of stocks, and
also, they have sufficiently well-diversified.
You should ensure that these, so that is where skill comes.
It's not mechanical application of Piotroski.
I want to re-emphasize again and again.
This is very important.
You can't just calculate.
And Excel also can do it, or a computer can do it.
You should exercise judgment here.
You should look at this logs and charts.
And you should ask yourself, is this sufficient data diversified?
Do I have enough number of stocks?
Am I exposing myself to some unforeseen risk?
So that is why I'm saying only nine may not be a good strategy.
If you have enough number of stocks, what do I mean by enough?
Some say 30, 35, 40 stocks with 10, 12 different industries,
I think you are fine.
You are doing okay, not bad.
Otherwise, go to eight, or even seven, for that matter.
Now that you've chosen your logs and charts portfolio, [INAUDIBLE],
that's the next question.
And how long to hold, that's another question.
And most importantly, when to close your trade.
What do I mean by close your trade?
Those which you have gone long, you will have to short.
And those you have shorted, you will have to long.
So that at the end of the period, you're back to zero.
And the difference between all these four prices is your profit.
Now, when, when to trade is the question.
First, let's answer that.
Can you guess, when is the earliest you can treat?
Pause for a while.
Of course, this medium does not allow me to wait for
your answer and then tell you, you are right, you are wrong, and all that.
That's not possible.
But I urge you to pause for a minute and think that,
when is that you can enter or start trading this strategy?
By when, I mean time of the year.
I guess you would have thought about it.
The earliest you can implement this strategy is
when financial statements are publicly made available.
Remember this.
None of our trading strategies are going to be based on private information.
It is not that you're an awesome director or some company official and
get it once information.
All our strategies are based on publicly disclosed information, and
therefore, you will have to wait until this information is implemented.
So even when you back test, this is very, very crucial, please note this.
Even when you back test, you're to pretend as if
you do not have information for a period, year end.
Until this results are disclosed,
you are to pretend as if you don't have this information.
What do I mean?
Let me give you a specific example.
Let's assume that your year ending in December 31st, right?
Now, you're doing a back test for, let's say, 2014.
Now, if you want to test that whether this particular strategy,
it doesn't necessarily need to be Piotroski,
it can be anything, has worked in the year 2014.
You cannot assume that on January 1st, you invested and held on for
three months, or six months, nine months.
Why?
Because on January 1st, you did not know the financial information.
Year end, board meeting happens.
Audit committee approves the numbers.
Then finally, the board approves the numbers.
And then they're disclosed.
And this process, depending on country to country, and
this is where institutional framework is important.
That's why I told you before that
you should have a look at that part of institutional framework.
And if you are trading in your home country, you will know all this.
A reasonably educated person will have an idea as to how much time it takes
between end of a period and disclosure of results.
So let's say we are in India right now.
So in India, it takes around two months, 60 days.
So even when you do your back test, you should factor this in.
And for the December example that I was talking about,
your trading should start on.
Say leave out January and February, say, March 1st.
So if your strategy is to buy and hold for three months, you start on March 1st, and
hold for another three months, another 90 days.
So this is very, very important.
So when do you buy?
It is immediately after public disclosure of information, right?
So again, if we summarize, we know how to score, we know what to select,
top and bottom, and all realities to buy, done.
Next, how long we should hold?
That's another question.
What should be the holding period?
Now, again, this will depend on your horizon.
Maybe you are like a not so active investor.
You may want to have a one year horizon.
Piotroski tests one year horizon in his paper.
Or you could want to try three months, or one month.
I would suggest not less than one month, because it's too short.
Because remember, these are stocks which are chronically undervalued by markets.
It will take time for markets to realize their value.
And hence, one month or 15 days may be too short a time.
One month is still okay.
15 days or a week, nothing is going to change in a week.
This strategy will,
again, almost guarantee you that one week holding period is not going to work.
So one month is the least I can think of, or 3, 6, 9, 12, anything is fine.
So decide on a holding period.
When you do the back test, test different holding periods.
So if you find that this strategy works only when you hold for 23 days,
should you be happy that you found this 23 day magic number?
Or should you start worrying?
Pause for awhile and think.
Let me rephrase my question if it is not that clear.
What I'm asking is suppose the strategy,
you do this back testing with different holding periods, 7 days, 15 days,
20 days, 1 month, 3 months, 6 months, so on and so forth.
Suppose you find that only a 23 day holding period works,
gives you a positive return.
No other holding period [INAUDIBLE].
Is it a good news or a bad news?
It is a big, big, big bad news.
There's something wrong.
There is nothing economically significant about 23 days.
Not you should walk only for 23 days.
This is where all this robustness tests come in.
Of course, there is a possibility if there is a peculiar institutional feature
applicable to that particular country,
where 23 days becomes important, then it's fine.
Then you should know what that institutional feature is.
Otherwise, the trading strategies should work for a range of this.
I'm fine with a strategy which works, lets say, between say 25 and
40 days, but slows down after 40 days and doesn't work beyond 6 months, lets say.
Because all this gets pricey.
That's fine.
But what will worry me is that it works for 40 days.
Suddenly on 41st day, it stops working.
That's not a good sign, right?
So try various time intervals and decide on your holding plan.
Once you're decided based on returns, so now you have the formula,
you have longs and shorts, you have your timing of investment, and
you are decided on your holding period.
So all elements are done.
Now, what is the next step?
No, no, no, no, not enough.
You cannot just start investing right now.
You'll have to wait further.
As I've suggested before, if you are a new trader, the first thing
that you should do once you are equipped with this formula is to do mock trading.
So there are number of platforms available in every country where you can
keep generating your buys and sells.
And using actual prices,
see how this trading strategy performs over a period of time.
Give it a couple of years.
There is no hurry.
See for a couple of years how this particular strategy works.
And two things might happen.
One is what are the two things in a trading strategy?
You make money, or you may not make money.
Fair enough.
But there is a third thing.
As you keep watching this, you may realize that some kind of tweaking.
Maybe you may realize that one signal is not informative at all.
You may think that once you [INAUDIBLE], ROI is not required.
So that is the purpose of this course.
You should be able to improvise.
You should be able to tweak.
You should be able to come out with a score keeping in spirit of Piotroski.
You should be able to come out with your own formula that works in your market.
So once you're confident that now you have a formula that works,
once you're confident that you've figured out what are the ingredients of your
formula, then only jump into that strategy.
Then only start implementing the strategy, and start buying and selling stocks.
So that's about Piotroski.
Thank you.
