I can prove "Random walk theory" wrong

[DMC31]

My research began back in 2005 which stemmed from an idea I saw when looking over stocks on the NYSE and NASDAQ markets. My brother (a database computer programmer) custom built a program to test out this idea.
The results certified to us, that stock market direction (from the opening price to the closing price of the day) could be predicted with high probabilities, incorporating this method.

The results also prove “Random walk theory” is wrong.

All trades back tested were, buying at the market open, setting a 4% stop loss, and selling on market close. No trades were ever held overnight.

Here are the results tested from January 1st 2002 – 27th February 2009.
Results:

2009: 45.75% return (33 days traded), (Probability of profitable days: 72.72%)

2008: 233.63% return (232 days traded), (Probability of profitable days: 67.24%)

2007: 96.29% return (208 days traded), (Probability of profitable days: 62.98%)

2006: 116.87% return (199 days traded), (Probability of profitable days: 68.84%)

2005: 121.14% return (205 days traded), (Probability of profitable days: 74.14%)

2004: 87.92% return (202 days traded), (Probability of profitable days: 65.32%)

2003: 159.21% return (209 days traded), (Probability of profitable days: 69.37%)

2002: 65.28% return (155 days traded), (Probability of profitable days: 52.25%)

Total return: 926.1% over 7 years and 2 months.

The results above are all long trades, when testing the system live, we were getting slippage on the short trades at market open, but when buying (long), we were guaranteed the opening days price. (When shorting stocks on market open, U.S markets requires a long trade before a short), that is why we were not getting filled at the market open price on short trades.

All stocks within the system trade 1-7 million in average volume per day and all trade on U.S exchanges.

Our method goes beyond being a high probability trading system with great returns, it proves random walk theory wrong, and certifies predicting stock market direction.

We contacted Google finance last week with a proposal, to prove our research, and are waiting for a reply.

Any feedback would be much appreciated on how to go about marketing our research and analysis,

 

[vincent191]

I too don't agree with the "Random Walk" theory either. However, those guys that came out with that theory did a very indepth study of the US market indices over a long period of time (from historic data) and they concluded that you cannot predict the movements in the indices with any degree of accuracy. Hence, the term Random walk....i.e. movements in the indices were at random.

I find that hard to accept because in my mind the stock market must react to some form of stimulus, either positive or negative. It cannot be at random.

 

[awg]

All trades back tested were, buying at the market open, setting a 4% stop loss, and selling on market close. No trades were ever held overnight.

Here are the results tested from January 1st 2002 – 27th February 2009.
Results:

2009: 45.75% return (33 days traded), (Probability of profitable days: 72.72%)

2008: 233.63% return (232 days traded), (Probability of profitable days: 67.24%)





The results above are all long trades,



when testing the system live, we were getting slippage on the short trades at market open, but when buying (long), we were guaranteed the opening days price. (When shorting stocks on market open, U.S markets requires a long trade before a short), that is why we were not getting filled at the market open price on short trades.

I am confused?

does the above imply you traded long only with the above quoted returns?

In your post you mention short trades, but as I read this, seems to mean that you only traded long.

to get such big returns over the period 08-09 without being short would be very surprising indeed

 

[Sean K]

Wow, where can I buy this?

 

[MS+Tradesim]

Wow, where can I buy this?

I was just wondering that too.

 

[dhukka]

Wow, where can I buy this?

me too, me too, where do I sign?

 

[DMC31]

The results I quoted are for all the long trades.

The return over the 7 years and 2 months including the short trades is 1205% with an average probability of 71.40% profitable days.

When we tested the system live we were not guaranteed the opening tick on the shorts at open, so we filtered out all of the short trades, leaving just the long trades. Removing the short trades, removed slippage.

 

[investorpaul]

Just wondering how did where you mention that you only went long and that you bought at open, did that mean you bought every day? or did you chose the days you bought on (if so, how)?

 

[Trembling Hand]

When we tested the system live we were not guaranteed the opening tick on the shorts at open, so we filtered out all of the short trades, leaving just the long trades. Removing the short trades, removed slippage.

Yeah yeah. We don't care about the details just take our payments for the system.

My Visa Card number is; 4586 0012 0007 0007 Exp 01/10 Just send me the black box NOW.

 

[MS+Tradesim]

Yeah yeah. We don't care about the details just take our payments for the system.

My Visa Card number is; 4586 0012 0007 0007 Exp 01/10 Just send me the black box NOW.

Woohoo!

*Orders Pizza and beer*

 

[James Austin]

live testing using your cash DMC31 and over an extended period before i get excited.

here's why . . . .

In 1997, two professors from Harvard won the Nobel Prize for Economics for services to easy money. They claimed to know pretty much everything, and weren't shy of sharing their knowledge for a fee.

All it took to make big profits, said Myron Scholes and Robert C. Merton, was a clever little formula. Oh, and a huge position in derivative contracts, gearing up tiny movements into large, volatile swings in options prices. Funnily enough, gearing up tiny movements into large, volatile swings in options prices was also all it took for Long-Term Capital Management to blow up in 1998.

Both of the Harvard laureates sat on the hedge fund's board of directors. LTCM lost some $3.5 billion.

 

[pilbara]

All trades back tested were, buying at the market open, setting a 4% stop loss, and selling on market close. No trades were ever held overnight.

how do you choose whether to go long or short?

did you use "market order on open" using real trades, or was this on demo? I'd expect significant slippage trying to get the opening price.

what typical brokerage commission was used?

 

[Sean K]

Woohoo!

*Orders Pizza and beer*

I've got a Ferrari on the way!

Cheers TH!

 

[Trembling Hand]

I've got a Ferrari on the way!

Cheers TH!

Ouch! whats happening??

 

[skc]

We contacted Google finance last week with a proposal, to prove our research, and are waiting for a reply.

Any feedback would be much appreciated on how to go about marketing our research and analysis,

Here's my feedback:

- Since when did Google Finance become the premier source of quantitative finance research?

- Being a first time poster showing ANY system on ASF is unlikely to generate much positive responses (this one included).

- If you only care about proving / dis-proving random walk, you don't actually need to place any trades and worry about slippage. You just need to show that your predictions are correct.

- If you are genuine in seeking feedback, then talk about what you want to achieve. Publish in major finance journal? Appear on next channel 7 "how to beat the sharemarket" special? or scam innocent forum posters?

 

[DMC31]

Setting a market on open order for long trades is easy and guarantees the open tick on Long trades (some brokers call them MOO orders), and closing out is (MOC) market on close order.

I won't disclose the idea behind the system, sorry, I have spent years researching this.

But I do want to prove it, that is why I contacted Google finance, and submitted a proposal, because I believe it goes beyond money making.

 

[James Austin]

Ouch! whats happening??

i'm not the only cynic, thank crikey!

 

[awg]

The results I quoted are for all the long trades.

The return over the 7 years and 2 months including the short trades is 1205% with an average probability of 71.40% profitable days.

When we tested the system live we were not guaranteed the opening tick on the shorts at open, so we filtered out all of the short trades, leaving just the long trades. Removing the short trades, removed slippage.

it indeed would be good to remove slippage

but because your system buys and sell within each day for every trade, seems to me you still would have a higher than average "trading system" slippage rate

 

[DMC31]

I chose to write a proposal to Google, because I believe they would be interested in more than just a trading system, it does prove some stock markets can be predicted,
I am not here to sell or endorse any company or any product, just my research.

 

[Frank D]

I personally think the random outcome has more to do with profits than expectancy.

In theory using back testing data the market rotates in UP and down far more often than it trends in 1 direction. So there is going to be a bias towards higher daily closes depending on the overall trend of the market

With a long bias of UP days it comes in around 71-72% since 1991 (without stops)

I’m sure the last year hasn’t helped with the current trend, but that will eventually turn around and the number of UP days will increase.

The only problem is the reward and profits associated with those UP days, which in my opinion becomes the random outcome in profits….

And obviously we don’t know exactly when those UP days will occur, but then again you can filter the days with stop losses or certain levels in the market.

Think about it…. How many higher closes in the past 4 weeks has the Aussie market had even though it keeps dropping.

12 out of 22 days have closed higher than the open…