patrick.net

adjusted for inflation no way 2000 can be beaten anyways. i think it is a programming/graphing error.

I blame Yahoos H1B programmers.

Wow, I've been seeing some anomalies on yahoo graphs myself, but had no time to dig into it. I agree with others, the quality of yahoo these days is crap. The speed of yahoo finance portfolios is abhorrent, too. It oftentimes takes 15 sec to refresh a portfolio.

But then, Google is also going downhill. Google has now reached the phase of a company where quality work does not count, and politicking is all that counts, by my estimation.

All the main services (*) of google are in decline. Gmail, Contacts, Maps, Drive, Voice, Calendar, FInance. No improvements, just new interfaces that often suck. Let me give some examples of lame stuff.

1. Gmail: they keep mucking with the interface, usually making it worse.

2. Contacts: The new "mobile" interface sucks

3. Maps: You cannot favorite ("mark with a star") the start and destination points when in directions mode. You have to back out into map mode, paste the address(es) in anew, and then you can do it one at a time. The printing of maps has sucked for years. Printing is far from WYSIWYG, problems with layout and general readability.

4. Drive: no write-only option for directories (useful when compiling files from multiple people without them having read or delete access). No way to upload from commandline to a directory by name, you have to have some hex link/name that has to be looked up one by one.

5. Voice: you can no longer click-to-call using obihai voip box as the callback destination. Amazingly lame and stupid.

6. Calendar: Add exception handling to recurring events, for example if it lands on a holiday
https://productforums.google.com/forum/#!topic/calendar/D0_whtUk3xk

7. Finance: They killed off much of the API, bad offering of ticker symbols covered, portfolios are not good, on and on.

justme says

But then, Google is also going downhill. Google has now reached the phase of a company where quality work does not count, and politicking is all that counts, by my estimation.

A friend of mine says everyone there is now "vesting in peace". Who would want to work very hard to make some of the richest people in the world even richer when you can just kick back and wait for your money?

justme says

It takes 15 sec to refresh a portfolio many times.

That's IF it refreshes or loads in the first place.

Patrick says

Very interesting that the max dot-com bubble peak you will see on Yahoo finance graphs for the Nasdaq is wrong, saying 3940.

IT has to do with time scale. Your graph show closing prices, i don't know monthly (Quarterly ?). Here, I'm going to post a screenshot of the yahoo daily graoh. Go to the one year graph and scroll back.

I noticed that yahoo only makes that mistake when you're zoomed out really far, showing many years of range(eg, at "max"). Try zooming out only 10 yrs, and then only 5 years - you'll see the peak keeps increasing.

I bet their programmer choose to use a time range averaging strategy (what was the average price for some 50 day time range? calc that, and show it as 1 dot on the zoomed out graph) , or maybe even just picking arbitrary data points, like the 7th of the month or whatever ends of being a midpoint of some time chunk, when choosing what to show when zoomed out that far.

Now that I think about it, I'm not sure how I would do it if I had too. Obviously, there's scenarios where you want to show the peak (this being a great example of that), but if you did that everywhere, I think it would lead to graph inflation, where when you zoom out, the whole graph looks a bit higher than it should be. Averaging would lead to a zoomed out graph thats more true to the typical values that existed, but you loose peaks and valleys.

Patrick - your graph shows one value (closing price) per quarter of a year. You can select bars that show open high low and close bars for each quarter. Hard to find though. Not under settings, just to the left of that.

I've looked into it more, and I think I was right - they use averaging.

Zoom out to max, and then move your mouse on the graph the very smallest amount you can to make the crosshairs move just 1 hop. For me, it jumps from Nov 1, '99 to Feb 1 '00 - a full 3 months. So, when zoomed out, they use a resolution of 3 months per data point, and then they just connect the dots w/ lines.

If I zoom to 5 years, each hop is only 7 days. So, the window used for calculating an average is nice and small, letting a peak/valley in the data influence the 1 wk average a lot.

This is kinda lame, they have plenty of pixels on the screen and could use a much better resolution than 3 months, albeit at the expense of needing the download more data points.

There still is truth out there Patrick. You'd be surprised. Even some of those Trump haters are coming from a place of truth and facts.

Interesting graph from a Wycoffian perspective.

c1561490 says

they have plenty of pixels on the screen and could use a much better resolution than 3 months

OR, you just look at a weekly or daily graph (or hourly etc). It's been done this way for as long as I can remember.