Registered User
Joined: 9/21/2005
Posts: 566
|
Interesting read from IBD about a talented guy. I still have my issues with the box method as a realistic form of duplicating his achievements (I think either key components were left out or maybe it just takes great acument to judge the breakouts), but you have to give credit where credit is due. I hadn't realized that he lived a relatively short life.
Nicolas Darvas was a renaissance man.
Literate as well as athletic, he was good at a wide range of activities.
He trained to be an economist at the University of Budapest.
He earned a living at occupations as disparate as creating crossword puzzles and sportswriting.
He played championship pingpong. He toured Europe and the U.S. as one of the world's highest paid ballroom dancers, book publisher Lyle Stuart noted in the preface of one of Darvas' works.
Darvas' most enduring feat was conquering Wall Street -- in his spare time. Still dancing full time, he parlayed a $3,000 bet on a speculative Canadian mining stock into a series of investments that culminated in a $2.25 million portfolio.
That success earned him a profile in Time magazine.
His 1960 book "How I Made $2,000,000 in The Stock Market" was a best-seller. It still pops up on reading lists for investors.
Darvas' book is notable for its skepticism about Wall Street's conventional wisdom.
He wrote in a later book titled "Wall Street: The Other Las Vegas": "Wall Street is nothing more than a huge gambling casino, bristling with dealers, croupiers and touts on one side and winners and suckers on the other. ... I had been a winner and was determined to stay one."
He stayed one through his discovery of a strategy based on patterns formed by stock share prices. It was a strategy that involved no recommendations from analysts or brokers, no hot tips, no financial stories.
Box Theory
The crux of his strategy was price and volume data.
Darvas (1920-77) developed what he called his box theory. As some stocks climb, their share prices stay within a range. Some break through the bottom of that range -- or box -- and are less likely to rally much soon. Others rise and set up a new, higher box.
As soon as a stock climbed to start a higher box, Darvas liked to buy.
Gradually, Darvas realized institutional support from big money on Wall Street helped keep top-performing stocks from falling out of their boxes.
What Darvas' strategy boiled down to was capitalizing on winning stocks' price gains.
That approach would be familiar to investors who use IBD's CAN SLIM technique and know the importance of charting stock patterns.
But his strategy was not nearly as comprehensive as CAN SLIM.
Darvas arrived at his technique through a mix of circumstance and trial and error.
Born in Hungary, he studied to be an economist at the University of Budapest. With the outbreak of World War II in 1939, he grew fearful of Nazis and communists vying for control of his country.
At age 23 with a forged visa and a little money, he fled to Turkey. Teaming with his half-sister Julia, they became one of the most popular professional dance duos in Europe and, later, the U.S.
In 1951 he came to the U.S. He trained diligently eight hours a day to hone his hoofer skills.
During downtime, he applied his dogged determination to learn about stocks. Soon he had read some 200 books on investing.
His first stock was offered to him in 1952 by two Toronto nightclub owners. The impresarios said they would pay him for a dancing engagement with 6,000 shares of Brilund, a Canadian mining firm. The stock then was worth 50 cents a share.
Darvas didn't take the Toronto gig. But to show goodwill, he bought their Brilund stock for $3,000.
He forgot about the stock until months later. He glanced at the stock's price in the paper. "I shot upright in my chair," he wrote in "How I Made." "My 50-cent Brilund stock was quoted at $1.90. I sold it at once and made a profit of close to $8,000."
Through a Canadian broker, Darvas bought 1,000 shares of a gold mining firm at $2.90. The following weeks it slid to $2.41. He sold.
Still, he began to buy a series of Canadian mining stocks. He acted on hunches, rumors, news about oil strikes. After seven months, he had lost about $3,000.
He tried to wean advice from advisory newsletters. He switched to a broker in New York. He tried more newsletters.
He plumbed the over-the-counter market for investments. When that didn't work, he switched to the Big Board. His hunt for winning stocks resembled hopscotch. He spotted companies with top earnings. He jumped to stocks paying big dividends. He leapt to industry leaders.
One was steel maker Jones & Laughlin. He raised money to buy shares by mortgaging his home and borrowing against an insurance policy. He asked the Latin Quarter nightclub in New York for an advance against his long-term contract.
He spent $52,652.30 for 1,000 shares.
But the stock dropped. He lost more than $9,000. "I was crushed, finished, destroyed," he wrote.
Depressed and facing bankruptcy, Darvas still spent hours daily studying newspaper stock tables. He spotted Texas Gulf Producing. He knew nothing about the company. All he saw was that its shares kept rising.
He bought 1,000 shares at 371/4. He watched it rise for five weeks. He sold at 431/4.
"The stock that saved me from disaster was one about which I knew nothing," he wrote. "I picked it for one reason only -- it seemed to be rising."
Darvas battled to learn from his mistakes with his usual persistence. He was resolute about not clinging to what he once thought should work -- but obviously did not.
After strenuous dance performances each night, he poured over stock tables. Now he ignored all of the factors he once thought were essential -- everything from newsletters to fundamentals.
"Clutching at a straw, I decided to make an extensive study of individual stock movements," he wrote. "How do they act? What are the characteristics of their behavior? Is there any pattern in their fluctuations?"
He read more books. He inspected charts. He noticed that as stocks moved up, they traded within a range. Leading stocks moved up again, trading in a higher range. He called each range the stock's box.
When a stock pulls back within a box -- but not falling out of it -- that is often a sign it is about to vault into a higher box.
Set To Spring
"Before a dancer leaps into the air he goes into a crouch to set himself for the spring," Darvas wrote.
Once Darvas spotted a stock ready to make its move, he could tell his broker to buy at a certain price.
He did all this no matter where he was. In New York, he studied stock prices at a favorite table in the popular Oak Bar of the Plaza Hotel. Abroad, he stopped at U.S. embassies to read newspaper stock tables.
Darvas had detractors. Dubious about the box theory, New York's attorney general in 1960 called Darvas' book fraudulent advice, but later settled out of court, according to the Boston Globe.
Other critics said Darvas was simply a shrewd stock picker who benefited from price momentum.
Darvas urged investors not to chase shortcuts. He wrote: "I have discovered no loss-free Nirvana. ... My method obviously wouldn't work for everyone. It worked for me. And, by studying what I did, I hope you find this book helpful and profitable for you."
|
|
Registered User
Joined: 10/7/2004
Posts: 2,126
|
you can find free darvas charting in the web. Specifically Tucows has it for free.
If you by any chance own Omnitrader software, they sell a Darvas module that you can add to Omnitrader.
I would give you the URLs, but I am sure the board coordinators will erase them.
Althought I don't use Darvas at all, I always found his story interesting.
good luck
|
|
Registered User
Joined: 9/21/2005
Posts: 566
|
Thanks BB. I dont know much about Tucows but will check it out. Dont care much for Omnitrader. I had tried it out a couple times but found its performance didnt match my expectations.
|
|
Registered User
Joined: 3/25/2005
Posts: 864
|
I haven't read "How I Made 2 Mil In The Stock Market" By Darvas.
But, I did read an overview about it once where (if I remember correctly), it was saying that the crux of his Box Theory is basically this:
1) Look for stocks making new 52-week highs.
2) Wait for them to retrace.
3) Get in when it looks like the retracement has stabilized.
4) Exit upon profits made as it reapproaches and/or surpasses the aforementioned 52-week high.
|
|
Registered User
Joined: 10/7/2004
Posts: 2,126
|
DARVAS CODE
//////////begin/////////
box1=0;
box2=0;
procedure fillDarvas(start,end,swap,top, bottom )
{
for ( j = start; j < end; j++)
{
if( box1[j] = swap)
box1[j]= top ;
else
box1[j]= bottom;
if(box2[j] = swap)
box2[j]= bottom ;
else
box2[j]= top;
}
}
BoxArr1 = 0;
BoxArr2 = 0;
StateArray = 0;
DBuy = 0;
DSell = 0;
TopArray = 0;
BotArray = 0;
tick=0;
BoxTop = High[0];
BoxBot = Low[0];
swap=0;
state = 0;
BoxStart = 0;
for (i=0; i<BarCount; i++)
{
if (state==5)
{
TopArray=BoxTop;
BotArray=BoxBot;
if (Low<(BoxBot*(1-tick/100)) || High>(BoxTop*(1+tick/100)))
{
fillDarvas(BoxStart,i,swap,BoxTop,BoxBot);
state = 1;
swap = !swap;
BoxTop = High;
BoxStart = i;
}
}
else
{
if (High<BoxTop)
{
if ((state<3) || (Low>BoxBot))
{
state++;
}
else
{
state=3;
}
if (state==3)
BoxBot=Low;
}
else
{
state=1;
BoxTop=High;
}
}
StateArray = state;
}
fillDarvas(BoxStart,BarCount,swap,BoxTop,BoxBot);
Plot(C,"",1,64);
//Plot( box1, "" , colorRed, styleLine );
//Plot( box2, "" , colorGreen , styleLine );
Plot( box1, "" , colorRed, 512);
Plot( box2, "" , colorGreen , 512);
///////////end///////////////////
jcfla7 this code is designed to be easily implemented in amibroker. If you do not use that plataform may be you can find a way to put this into yours.
good luck
|
|
Registered User
Joined: 10/7/2004
Posts: 426
|
Your coding is not correct. Here is the correct code.
//////////begin/////////
box1=0;
box2=0;
SetBarsRequired(10000,10000);
procedure fillDarvas(start,end,swap,top, bottom )
{
for ( j = start; j < end; j++)
{
if( box1[j] == swap)
box1[j]= top ;
else
box1[j]= bottom;
if(box2[j] == swap)
box2[j]= bottom ;
else
box2[j]= top;
}
}
BoxArr1 = 0;
BoxArr2 = 0;
StateArray = 0;
DBuy = 0;
DSell = 0;
TopArray = 0;
BotArray = 0;
tick=0;
BoxTop = High[0];
BoxBot = Low[0];
swap=0;
state = 0;
BoxStart = 0;
for (i=0; i<BarCount; i++)
{
if (state==5)
{
TopArray=BoxTop;
BotArray=BoxBot;
if (Low<(BoxBot*(1-tick/100)) || High>(BoxTop*(1+tick/100)))
{
fillDarvas(BoxStart,i,swap,BoxTop,BoxBot);
state = 1;
swap = !swap;
BoxTop = High;
BoxStart = i;
}
}
else
{
if (High<BoxTop)
{
if ((state<3) || (Low>BoxBot))
{
state++;
}
else
{
state=3;
}
if (state==3)
BoxBot=Low;
}
else
{
state=1;
BoxTop=High;
}
}
StateArray = state;
}
fillDarvas(BoxStart,BarCount,swap,BoxTop,BoxBot);
Plot(C,"",1,64);
Plot( box1, "" , 1 + statearray, styleLine );
Plot( box2, "" , 1 + statearray , styleLine );
|
|
Registered User
Joined: 10/7/2004
Posts: 426
|
For some reason or another the i inside of the brackets do not post correctly.
|
|
Registered User
Joined: 10/7/2004
Posts: 2,126
|
the code is perfectly fine. Are you trying this in amibroker.
Make sure to do an inser and not a drop on chart - you cannot apply to existent chart panel.
|
|
Registered User
Joined: 10/7/2004
Posts: 426
|
The code is not fine. cut and paste in amibroker and click on the AFL button and see the errors. Some lines are missing the 'left bracket' i 'right bracket'
|
|
Registered User
Joined: 10/7/2004
Posts: 2,126
|
Ok Booker I don't have a lot of time to give you input on this, but I will show you that is working fine, and I am going to give you further code to do it in a different way.
Something must be different with your settings or the way you are trying to implement.
Here let me show you.
Right below is the code I gave you originally.
and below the results when implementing in my system
HERE below is a code to do basically the same a little different
and here below is the result
That is as far as I will go on this. Good luck.
|
|
Registered User
Joined: 3/25/2005
Posts: 864
|
Oh, Ok.
I think I see how it works now:
1) Look For the next 52 week high (Box Top).
2) Allow price to retrace to form Box Bottom (as will be indicated just before the next higher low and higher high on the next candle occurs).
3) Look for the next 52 week high to occur after the one formed in #1 above (this one breaches the Box Top formed in #1 above).
4) Allow price to retrace.
5) Get in after the first higher low and higher high occurs (indicating the retracement has finished).
6) Exit upon profits made on the way up on the currently rising leg.
|
|
Registered User
Joined: 3/25/2005
Posts: 864
|
Can a PCF be created from the above described scenario?
|
|
Registered User
Joined: 9/21/2005
Posts: 566
|
Thats a good question - was thinking the same thing. Also, I thought Darvas had way of riding some of his long picks for months? And didnt he add to his positions?
Hohandy, if you read this - didn't you study Darvas?
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Nope, not me - this thread is the first I've heard of Darvas. Might be thinking of Tommy...
|
|
Registered User
Joined: 9/21/2005
Posts: 566
|
My bad, it was fpetry that I chatted about Darvas with.
Here is the link (before Diceman beats me to it)
http://www.worden.com/training/default.aspx?g=posts&t=21008
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
Thanks for the reminder JC - I do remember that thread now, and I remember thinking at the time that it might be an interesting read - and then promptly forgot about it...
I was given the Turtle book recently by my secretary, who's husband is a professional trader and has some connection to the Turtles - so I promised her that I'd read that next (I'm much more afraid of my secretary than I am of my boss LOL!)
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
I don't exactly remember what the Darvis system
was. (just that it involved breakouts)
(I read the book in the mid 1980's)
I know that in the last few years Guppy did some
original work on this.
(if I recall a 100 day high was the box entry)
(He did some articles for Technical Analysis
of Stocks and Commodities magazine)
-------------------------------------------------------------
I also heard the system described as:
If a stock fails to make a new high for 3 days,
the most recent high that is higher than the three
subsequent highs becomes the box top.
A chart was shown with AAPL that showed a
buy on about 7/20/06 that sold on 10/10/06.
The next buy was 11/09/06 and sold on 12/12/06.
From this I do not believe a 52 week high is
necessary.
(since these are under the 1/17/06 high)
I think the boxes self-adjust as you go.
Thanks
diceman
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (memorableproducts) I haven't read "How I Made 2 Mil In The Stock Market" By Darvas.
But, I did read an overview about it once where (if I remember correctly), it was saying that the crux of his Box Theory is basically this:
1) Look for stocks making new 52-week highs.
2) Wait for them to retrace.
3) Get in when it looks like the retracement has stabilized.
4) Exit upon profits made as it reapproaches and/or surpasses the aforementioned 52-week high.
If he'd only had the Diceman indicator!
|
|
Registered User
Joined: 10/7/2004
Posts: 426
|
Ok BigBlock this is the last time I will try to explain it to you. for example in your original post look for the line TopArray=BoxTop. Do you see something missing? look real hard. Can you see that you cannot have an array without the "left bracket" "i" "right bracket". Now look for that same line in your posted picture and you will see the missing text. If you have time, look to see if the rest of the loop is missing the "i" array. I am not saying you have erroneous code on your computer, I'm telling you that it did not post correctly.
|
|
Registered User
Joined: 3/25/2005
Posts: 864
|
Booker,
I see your point.
The variable(s) must have been declared globally to the application.
I don't see "j" declared either.
I don't know anything about amibroker but if this is what you are using then you will need to figure out how to post any undeclared variables, globally, in order to get the code to run.
Or ask BB to send you a copy of the global variables section.
|
|
Registered User
Joined: 10/7/2004
Posts: 2,126
|
booker just take the code from the pic. It is possible I copied originally a draft or else. Sometimes I am working at lightning speed.
The pic code works as you see.
Take care
|
|
Registered User
Joined: 10/7/2004
Posts: 426
|
BigBlock I do not need this code. I already have the correct code and about 1,000 more I was just pointing out that some of the posting was dropped. It is not something you did, there is a problem on this site that removes text from some posts. If you look at the code I posted under yours you will see the same thing happened to me. That was the code from my file.
|
|
Registered User
Joined: 10/7/2004
Posts: 2,126
|
Well, that is nice to know. So you are saying that the board coordinators are changing the code. I see no other way this would happen.
|
|
Registered User
Joined: 12/21/2004
Posts: 902
|
QUOTE (BigBlock) Well, that is nice to know. So you are saying that the board coordinators are changing the code. I see no other way this would happen.
That's not it at all - certain symbols when typed within brackets get confused with HTML language and doesn't show up because it isn't interpreted by the software program running this site as text that is supposed to be read as "text". I've occaisionally run into this problem on other sites also
|
|
|
Guest-1 |