I've noticed this strange anomaly a few times now and the most recent example is in SMH. if you look at any chart provider you'll see the recent consolidation lows at around 27.30ish. Only on Tuesday did it break 27 to open around 26. Finviz however shows a different story. They show the recent consolidation lows to be around 26, and yesterday it opened at 26 as well so there is a very clear support line at 26. I don't understand how finviz quotes can be so far off. we're around over a dollar difference not a penny or two as one might expect. I've attached some charts below for comparison. Please put forth your theories as to why this might be. This really makes me loose confidence in Finviz as i use them all the time to screen for support/resistance in stocks/etfs. <img src="http://chart.finance.yahoo.com/z?s=SMH&t=6m&q=c&l=on&z=l&a=v&p=s&lang=en-US®ion=US"> <img src="http://finviz.com/chart.ashx?t=SMH&ty=c&ta=1&p=d&s=l">
This is very interesting. It looks like the price scale is two low by about $1-1.5 for that whole period? Have you feed that back to FinViz? I'm intrigued now. It doesn't look to be a dividend adjustment problem or a split adjustment problem but it could be I guess... You say you've seen this often?
Sent them an email and got: "Already replied several times on this issues today. The FINVIZ charts are dividend adjusted just like stockcharts.com" I compared the close and the adjusted close from yahoo historical and does seem to fit. Yahoo! does not use the adjusted close in the charts but provides it in the historical data. Looking at Google Finance, I do see several dividends paid in the last year from the constituents of the ETF so.
thanks the responses. I looked at stockcharts and now they have adjusted their charts. it seems that there is a delay for that to occur. So this presents a pretty big issue because how can one judge real support and resistance levels when people are looking at false tops and bottoms? That is to say people look at a chart and say well smh has support at 26.00 whereas the people that actually traded the stock actually were buying at the real support area of 27.50 not 26. This screws everything up.
Historical data is always tricky. For decision making, it may be useful to look at non adjusted data, since on the tape, it may be perceived as a price fluctuation. But when dealing with performance and returns, it's important to adjust because from a stock holder perspective, holding 100@28 or 100@27.50 + $50 cash (from a .50/per dividend) is the same. Following the dividend, your base price gets adjusted and every price point in between. It's something backtest systems have to wrestle with constantly. Most data providers offer both. When downloading Yahoo! historical for example, it contains a 'close' and an 'adjusted close'.
I see, well i'm glad i'm aware of this now, I can't believe it took me 2 years to figure this out. I used to always think daily charts never included dividends. And yes it makes sense from a long term perspective to include dividends in the chart. However, now i gotta take look at the chart without dividends in order to figure out the true psychological support/resistance levels.