EOD - Data Bull - HQuotes

[guibraga]

Hello Everybody !
I´m in Brazil and I want to start follow US Stocks End-of-Day.

Does anyone know about the quality of the Yahoo Data (where Data Bull and HQuotes get it) ???

If the data is not good... where can I get free quality EOD Data ?

Thanks a lot !
Guilherme

 

[epsgrowth]

Hquotes add in dividends to price figures

I have been trying to find a good program and data feed and thought that hquotes was a good selection; however, I found out after I bought it that the figures they provide add in dividends. The adjust for splits, but because of the dividend problem, I can't use their data.

I have written to them for help, but after sending 5 messages, I have not gotten even 1 reply. Turns out they are in Russia so I am not sure how receptive they will be.

I am looking for a source that provides 10+ years of data and adjusts for splits and doesn't add in dividends as though you reinvested the dividends. Or better yet, it would be nice if they gave you the option to adjust for splits and dividends.

Thanks

Hello Everybody !
I´m in Brazil and I want to start follow US Stocks End-of-Day.

Does anyone know about the quality of the Yahoo Data (where Data Bull and HQuotes get it) ???

If the data is not good... where can I get free quality EOD Data ?

Thanks a lot !
Guilherme

 

[RichardDale]

There are various Yahoo downloader programs out there. Some are free, some cost money. However, as you've already determined there's a significant amount of problems with the data including:

1. No delisted history - any system testing you do based on Yahoo data is subject to population bias. ie you can only test on the stocks that are currently listed.

2. Significant inaccuracies for any stock exchange that trades in increments less than 1c. This includes the Australian Stock Exchange (ASX), Singapore Exchange, some NASDAQ stocks and most OTCBB & Pink Sheet stocks. This happens because Yahoo's data is rounded to the nearest cent. For example, there are 14000 listed instruments on the ASX. Over 8000 of them are currently trading below 20c which means they move in either 0.5 or 0.1c increments. Since Yahoo only prices in cents this introduce VERY SIGNIFICANT innacuracies. - eg. a 2.6c shown as 3c represents a 15% error).

3. Normal dividends are adjusted by subtracting all historical data by the dividend amount. Exchanges such as the ASX do not adjust price data for standard dividends. By diluting it the way they do, any technical analysis or backtesting no longer shows a correct % return, nor does it show a critical technical levels such as support and resistance.

4. No sector information - critical if you want to perform sector-based analysis.

5. Significant number of missing stock splits and other capital adjustments.

6. There is no guarantee that Yahoo will continue to supply data, especially to automated download programs. In Yahoo's Terms of Service, they have a clause which says "You agree not to reproduce, duplicate, copy, sell, trade, resell or exploit for any commercial purposes, any portion of the Service (including your Yahoo! I.D.), use of the Service, or access to the Service.". Companies that produce Yahoo downloaders are effectively exploiting the Yahoo service and are therefore in breach of the Yahoo Terms of Service and Yahoo may shut down these types of automated Yahoo downloading programs very easily.

So, whilst the Yahoo data is free to obtain from the Yahoo site, you need to determine what you really need from the data. If you're every likely to want to view historical charts (eg for system development / backtesting / historical interest) then the old adage "you get what you pay for" certainly applies here.

The best analogy I can use for data feeds is they are like a game of Chinese Whispers. http://www.chinesewhispers.com/chinesewhispers.php

In the case of programs that extract data from Yahoo, the exchange sends data to Reuters. Reuters interpret the data, put it into their databases, then take an extract of the data and send it to Yahoo. Yahoo interprets the data, puts it into their database, then extracts it onto the Yahoo Finance web site. Then the downloading program obtains data from the Yahoo site, interprets it, and puts it onto the user's PC. If at any step of the above process, the interpretation is slightly incorrect then you have a discrepancy between actual events on the market and data you receive. This is clearly the case with the Databull/HQuotes/Quotes4U products on the market.

The data product I'm responsible for (Premium Data) is not adjusted for cash dividends but is adjusted for other items (such as splits, consolidations, demergers, capital returns etc.)

Cheers,
Richard.

 

[epsgrowth]

Thanks for the info. I was interested to know why some services adjust for dividends? When you look at stock charts, are they adjusted for dividends? I don't think they are, but I have never really thought about it much. How about Standard and Poors? When they list the hi and lo stock prices for each year, have they been adjusted for dividends? I know they are adjusted for Splits..

I also looked at premium data. I am a bit confused. It shows a one time fee of $90 for historical data. Is that the data that is adjusted for splits but not for dividends? Also, what is the other EOD cost? What do I get with EOD that I don't get with Historical data? I want hi, lo and closing prices for the stocks I cover.

Thanks,

There are various Yahoo downloader programs out there. Some are free, some cost money. However, as you've already determined there's a significant amount of problems with the data including:

1. No delisted history - any system testing you do based on Yahoo data is subject to population bias. ie you can only test on the stocks that are currently listed.

2. Significant inaccuracies for any stock exchange that trades in increments less than 1c. This includes the Australian Stock Exchange (ASX), Singapore Exchange, some NASDAQ stocks and most OTCBB & Pink Sheet stocks. This happens because Yahoo's data is rounded to the nearest cent. For example, there are 14000 listed instruments on the ASX. Over 8000 of them are currently trading below 20c which means they move in either 0.5 or 0.1c increments. Since Yahoo only prices in cents this introduce VERY SIGNIFICANT innacuracies. - eg. a 2.6c shown as 3c represents a 15% error).

3. Normal dividends are adjusted by subtracting all historical data by the dividend amount. Exchanges such as the ASX do not adjust price data for standard dividends. By diluting it the way they do, any technical analysis or backtesting no longer shows a correct % return, nor does it show a critical technical levels such as support and resistance.

4. No sector information - critical if you want to perform sector-based analysis.

5. Significant number of missing stock splits and other capital adjustments.

6. There is no guarantee that Yahoo will continue to supply data, especially to automated download programs. In Yahoo's Terms of Service, they have a clause which says "You agree not to reproduce, duplicate, copy, sell, trade, resell or exploit for any commercial purposes, any portion of the Service (including your Yahoo! I.D.), use of the Service, or access to the Service.". Companies that produce Yahoo downloaders are effectively exploiting the Yahoo service and are therefore in breach of the Yahoo Terms of Service and Yahoo may shut down these types of automated Yahoo downloading programs very easily.

So, whilst the Yahoo data is free to obtain from the Yahoo site, you need to determine what you really need from the data. If you're every likely to want to view historical charts (eg for system development / backtesting / historical interest) then the old adage "you get what you pay for" certainly applies here.

The best analogy I can use for data feeds is they are like a game of Chinese Whispers. http://www.chinesewhispers.com/chinesewhispers.php

In the case of programs that extract data from Yahoo, the exchange sends data to Reuters. Reuters interpret the data, put it into their databases, then take an extract of the data and send it to Yahoo. Yahoo interprets the data, puts it into their database, then extracts it onto the Yahoo Finance web site. Then the downloading program obtains data from the Yahoo site, interprets it, and puts it onto the user's PC. If at any step of the above process, the interpretation is slightly incorrect then you have a discrepancy between actual events on the market and data you receive. This is clearly the case with the Databull/HQuotes/Quotes4U products on the market.

The data product I'm responsible for (Premium Data) is not adjusted for cash dividends but is adjusted for other items (such as splits, consolidations, demergers, capital returns etc.)

Cheers,
Richard.

 

[RichardDale]

There are three methods for making adjustments for cash dividends.

1. Don't adjust at all.
If this is the case then all of your charts will show accurate levels of where real trades actually occured (eg. for support/resistance/trend lines). The theory is that people holding stocks only factor in the prices they actually paid for the stock, not the dividends they may have received along the way.

2. Adjust proportionally.
Here's an example:
Day 1: Stock trades at $10.00
Day 2: Stock trades at $10.50
Day 3: Stock trades at $11.25
Day 4: Stock goes ex-dividend, the dividend is $0.75, and finishes trading at $10.50

If we adjust proportionally the data becomes:
Day 1: $9.33
Day 2: $9.80
Day 3: $10.50
Day 4: $10.50 (the ex-day is not adjusted)

Benefits:
The % return from holding on Day 1 through Day 4 is retained and incporates any dividends paid.

Disadvantages:
The prices do not represent reality. Many technical analysts like to view data without adjustment for dividends as market behaviour is not really affected by dividends (eg. support & resistance levels).

3. Back adjust by a fixed level.
This is the method used by Yahoo. Here's the same example:
Day 1: Stock trades at $10.00
Day 2: Stock trades at $10.50
Day 3: Stock trades at $11.25
Day 4: Stock goes ex-dividend, the dividend is $0.75, and finishes trading at $10.50

Now, if we back adjust the data becomes:
Day 1: $9.25
Day 2: $9.75
Day 3: $10.50
Day 4: $10.50 (the ex-day is not adjusted)

Benefits:
The actual $ return from holding on Day 1 through Day 4 is retained and incporates any dividends paid.

Disadvantages:
- The prices do not represent reality. Many technical analysts like to view data without adjustment for dividends as market behaviour is not really affected by dividends (eg. support & resistance levels).
- % return is incorrect.
- Any indicators (such as moving averages, RSI etc.) will trigger at different levels than the previously unadjusted data.

----------------------------------​

Because of the disadvantages of methods 2 and 3, we have decided NOT to adjust for cash dividends. Our data is, however, adjusted for any capital events (bonus issue, stock split, consolidation, capital returns, demergers etc.).

Regarding the Premium Data pricing, we charge a fixed price for historical data. If you want daily end-of-day updates then you can also pay for that too. We offer a free trial so you try-before-you-buy.

Cheers,
Richard.