I gave up on Google long ago. Their data is spotty and error-prone. Plus, the quotes are unadjusted for dividends. I had been looking at them because I was trying to decide if I wanted to do some specific functions for them using my EXCEL add-in.
I’ve looked at them several times over the years. Last time I looked, I found a number of days with 0 volume, where O/H/L/C was set to the closing price of the day before.
Just checked SPY. The more recent data has no zero volume days, but 2012-01-13, 2011-12-30, and 2011-12-11 all do. Going back to 2001, there are 19 days with zero volume. So it looks like they may have improved a bit.
And, just going back to the start of 2014, they show a closing price of $182.92. The adjusted closing price should be about $171.35.
Their option data is even worse – missing expiration dates and contracts left and right. They don’t even have an “Option chain” link for SPY, and MMM only shows a single expiration date, while there are actually 12 of them.
For 2016 and 2017, I see three times where closing price differs from Yahoo and 99 times where volume differs. For example, 2/29/16 shows $193.35 vs $193.56 and 122448542 vs 125918100. Adjusted close should be $188.67. MarketWatch shows the Yahoo values, but it could just be that MarketWatch uses the same historical quotes provider.
BTW, for my add-in processing, I am creating the adjusted closing prices myself, using the dividend stream shown on Yahoo. They’re supposed to be fixing the data for dividend adjustments, but I won’t hold my breath. I needed to find an alternate way to grab prices from Yahoo after they obsoleted their old CSV API. If they ever fix the dividend adjustment, I’ll just unplug my dividend adjustment calculation.
I should mention Yahoo data is spotty right now as well. Indexes and ETFs appear to have days where they have zero or null values. But Yahoo does seem to be scrubbing the data. I also found a case, RNN, where they applied a 1-for-10 split as if it were a 10-for-1 split.
But I’m not sure we should even call it “Yahoo” or “Google” data. They must have a provider for their data. I know at one time, Yahoo charts did not agree with their historical quotes, because the data came from different providers – the charts were based on a provider that didn’t adjust for dividends. Well, some adjustments were made, but not all. 
For those of you that claim dividend adjustments are unnecessary, do you really want a dividend payment to trigger a technical buy or sell signal? Sure, dividends are usually small, but ignoring them is like ignoring stock splits. After all, stock splits don’t happen very often, right? I recall in 2008 when a lot of ETFs and funds paid out large distributions at year end, some more than 10%, there were a lot of complaints about big shifts in something as simple as a moving average.
For example, SDS had a $46 distribution in December of 2008 when the price was $350, about 13%. They also paid out about $16 three months earlier.
