One Page Yahoo Wonderithm

[ThirtysixD]

I recently made a bet with a family member who happens to be very high up in the big 4.

He seems to think that investing takes a huge amount of skill whereas I like to argue that that anyone that can follow a simple model will do fairly well for themselves. After seeing the other posts on this sub-forum I felt this would be a good place to keep tabs on the progress of the bet.

Without going into too many boring details the bet comes down to me being able to come up with a system that comfortably beats the market using only one page of yahoo data.

The rules for my algorithm are really simple:

1. All data must come from the yahoo page found at 'https://au.finance.yahoo.com/q/ks?s'
2. Companies are ranked by several factors with the top 20 being purchased each month
3. Companies must have positive operating cash flow & return on equity
4. The 50 day simple moving average must be above the 200 day
5. No stop loss or risk control

I will be posting an updated spreadsheets on this page at the beginning and end of each month to show performance.

Here is a backtest on US stocks showing how the strategy did from 2005 until present day.
Sharpe = 1.87, Sortino = 2.51, Drawdown = 59.9%


There are no guarantees for future performance but if I can make one snobby investment banker slightly less conceited then the world will be a much nicer place for everyone

*Disclaimer intended for entertainment purposes only
**Do not consider anything on this page as investment advice

 

[ThirtysixD]

First months stocks

Stock
HIT.AX
NST.AX
RMS.AX
FRM.AX
DTL.AX
SBM.AX
DRM.AX
KAM.AX
CII.AX
ELD.AX
HHl.AX
SIV.AX
QAN.AX
EVN.AX
BWF.AX
KKT.AX
FID.AX
SIQ.AX
AAP.AX
RRL.AX


If anyone can share an easy way to post tables on here that would be great

 

[CanOz]

If one of the factors is momentum, even in termso f % gain over a recent period...then this sound remarkably like one of Radge's systems....

 

[ThirtysixD]

If one of the factors is momentum, even in termso f % gain over a recent period...then this sound remarkably like one of Radge's systems....

Cant say I have read any of his books yet but yep it uses momentum & fundamentals. Its really just another magic formula variation with momentum (nothing too propriety as the yahoo page is somewhat limited).

 

[skc]

AAP.AX

AAP is a shell. May be a simple market cap >$10m filter to eliminate them?

If anyone can share an easy way to post tables on here that would be great

I normally just put numbers in Excel and post a screen capture.

The rules for my algorithm are really simple:

1. All data must come from the yahoo page found at 'https://au.finance.yahoo.com/q/ks?s'
2. Companies are ranked by several factors with the top 20 being purchased each month
3. Companies must have positive operating cash flow & return on equity
4. The 50 day simple moving average must be above the 200 day
5. No stop loss or risk control

I will be posting an updated spreadsheets on this page at the beginning and end of each month to show performance.

Interesting wager.

Can you explain the buying / selling rules?
- Assuming all positions of equal size?
- Do you simply buy a fresh batch of 20 stocks per month (unless the scan comes up with the same stocks)?
- Do you sell when the MA50 crosses under MA200?

 

[Gringotts Bank]

36DD, I'm imagining a fair bit of survivorship bias with this approach, but hard to know without the rules and without knowing the average hold time.

 

[CanOz]

Apparently there is some good research around about similar momentum style ranking systems. I like the index filter, if as SKC says, it exits you to cash at some point.

Your back tests, they include fundamentals as well, hows that work?

Anyway, good luck on the wager. The research that I've seen is that its a solid approach (momentum investing) that can only be improved by adding the fundamental piece.

 

[minwa]

Goodluck hope you win ! Not surprised someone high on the corporate ladder would think that way. Corporate environment is the most brainwashed environment, beside religion and North Korea. They are are of course programmed to believe you can't beat the market so just chuck it into funds with massive fees.

Anyway interesting to note you back tested on US stocks but are trading ASX. Any reason for this ? Does ASX data not have enough to be back tested ?

 

[ThirtysixD]

AAP is a shell. May be a simple market cap >$10m filter to eliminate them?

I normally just put numbers in Excel and post a screen capture.

Interesting wager.

Can you explain the buying / selling rules?
- Assuming all positions of equal size?
- Do you simply buy a fresh batch of 20 stocks per month (unless the scan comes up with the same stocks)?
- Do you sell when the MA50 crosses under MA200?

I thought about excluding AAP because of its tiny market cap but it is a real company.

Yes its simply the top 20 each each month (however if only 15 pass through the sma filter then parcel size will be increased).
No selling until the end of the month (you would add some risk control in a real system but this would require more price data).
Purchase price will be the price of the first trade of the month (2nd for this month because I was not fully organised).

 

[ThirtysixD]

Goodluck hope you win ! Not surprised someone high on the corporate ladder would think that way. Corporate environment is the most brainwashed environment, beside religion and North Korea. They are are of course programmed to believe you can't beat the market so just chuck it into funds with massive fees.

Anyway interesting to note you back tested on US stocks but are trading ASX. Any reason for this ? Does ASX data not have enough to be back tested ?

I have my own asx database but it is not very long compared to what Quantopian offers for US stocks (I chose to show a backtest over a decent period). Both markets are similar but I think fundamentals don't matter as much here (there is a stronger momentum effect). You can buy asx data for prices that only institutions can afford.

Apparently there is some good research around about similar momentum style ranking systems. I like the index filter, if as SKC says, it exits you to cash at some point.

Your back tests, they include fundamentals as well, hows that work?

Anyway, good luck on the wager. The research that I've seen is that its a solid approach (momentum investing) that can only be improved by adding the fundamental piece.

Historic EV/EBITDA and ROE data are used too.

I am using momentum & volatility plus fundamentals in my real accounts and it seems to be nice and robust. But like all good things it will stop working once the space gets crowded.

36DD, I'm imagining a fair bit of survivorship bias with this approach, but hard to know without the rules and without knowing the average hold time.

You can prevent most survivorship problems by properly adjusting your data (i.e dont let it see into the future). Overfitting is always a problem but theres not much you can do other than keeping things simple and rationale.

 

[ThirtysixD]

Mid month update >>


I just realized that Yahoo rounds price data.

AAP didn't move in price but ill need to add 10c minimum price filter for next months stocks.
WRR was the 21st ranking stock so returns would have been slightly higher.

The dumb criteria is also outperforming my more complicated factor model (live account).