Time series (indices, macro data and SP500 estimates) are now more easily accessible. Also new FRED series available to rules.

Dear All,

We cleaned up things related to accessing and using times series. The main changes are the following:

  • You no longer need to use GetSeries, like Close(0,GetSeries(“$VIX”)). You can now simply say: Close(0,$VIX). Both ways are supported.
  • Several FRED series were added like 3Mo T-bill, Prime rate, etc.
  • Reorganized the Rule Reference

More FRED series will be added to match the Macro Charts. Let us know if there others that need to be added.

Thanks.

Here’s a list of all changes related to series:

  • All series id’s are now found in the folder TIME SERIES in the rule reference
  • The frequency of series varies (like daily, weekly,etc). You will find that in the description of each series
  • Examples have been added in the Rule reference
  • We added several macro series from FRED. We only added FRED series that are not revised.
  • We moved all economic data from Compustat to a folder “Compustat Macro - TO BE DEPRECATED”. Please try to transition to other series
  • We will soon add access to the series that have revisions (like Unemployement) within rules.
  • We will also be deprecating #TNX, a strange daily time series of the 10Y Note that needs to be divided by 10 to get the conventional rate (yesterday’s value is 23.31). If you use #TNX, please use ##UST10YR. If you are not sure what to do, let us know.

Hi Marco,

you totally rock!!!

It would be great to have your Breadth Indicators (%AboveSMA200-SP500, …)
Thank you very much,

Toby

Thanks Marco!

This is great news! Having more data and shorter update intervals will take our market timing efforts to a new level.

Here are some additional suggestions:

  • Total Non-Farm Employment: this is currently available as Compustat #EMPLOY and it’s one of the most effective market timing indicators
  • Crude Oil Prices (OIL): Very effective market timing indicator (most bear markets were preceeded by a spike in oil prices)
  • SP500 Shiller Earning (SPEARN)
  • Real Personal Consumption Expenditure (RPCE)
  • Wages in manufacturing
  • Capacity Utilization (CAPUTIL)

Best Regards

I am using #UNEMP and #HOUSE and it would really annoy me if they are not replaced by another source.
Using moving averages limit the bias of revisions.
When will series “to be deprecated” be discontinued?

Edit:
if we could get the VIX:VXV ratio, it may be interesting

Fred, We won’t deprecate them until we have the FRED point-in-time replacement. How long is your moving average period ?

Thanks Marco!

Two thumbs up!!!

Between 3 and 6 months.

Thank you for simplifying things Marco.

Thank you so much for your continuing work on p123’s features and data. It’s really great.

One question: Shouldn’t #TNX and ##UST10YR be the same? I’m not sure that’s the case.

To be more specific…

The most recent value of #tnx shows 23.93 while ##ust10yr shows 2.35. After making the 10x adjustment, they are not the same.

Likewise if you check values on July 1st, #tnx was 24.18 and ##ust10yr was 2.35.

Does transitioning from #tnx to ##ust10yr have consequence for the calculation of #sprp?

Thank you.

For July 1:
Yahoo Finance: CBOE Interest Rate 10 Year T No (^TNX) close = 2.42%
FRED: 10-Year Treasury Constant Maturity Rate = 2.43%
U.S. Department of the Treasury: 10-yr Daily Treasury Yield Curve Rate= 2.43

For June30:
For July 1:
Yahoo Finance: CBOE Interest Rate 10 Year T No (^TNX) close = 2.34%
FRED: 10-Year Treasury Constant Maturity Rate = 2.35%
U.S. Department of the Treasury: 10-yr Daily Treasury Yield Curve Rate= 2.35

Thank you, geov. I can see that #tnx is the 10 year T-note quote and ##ust10yr is the 10 year constant maturity. They are of course similar but not the same. Both series are valuable.

Here is the data in p123’s database for the two variables during the last two weeks:

date #tnx ##ust10yr
6/22 2.360 2.370
6/23 2.409 2.420
6/24 2.371 2.380
6/25 2.393 2.400
6/26 2.476 2.290
6/29 2.331 2.330
6/30 2.335 2.350
7/1 2.418 2.350
7/2 2.393 2.350

Please note also that ##ust10yr appears to be rounded while #tnx is not.

Finally, ##ust10yr does not seem to have updated since 6/30. Is that an aberration or will ##ust10yr “run late”?

Related, will the transition from #tnx to ##ust10yr have consequence for the calculation of #sprp?

Thanks very much!

I’m sorry… The data points on 6/26 should be:

6/26 2.476 2.490

Any update on the addition of new FRED series for rules? Or perhaps simply allow a rule to access a ‘Rank of Binary Signals’ series from a particular macro chart?

ALFRED (historical FRED) series are point in time. Only our MACRO CHART can handle them for now. The MACRO chart is it’s own , self contained software module. We need to add support in the simulation engines.

It’s on the to-do list but it has been pushed aside several times because simply adding it to the current sim engines would only simulate from 1999. However we have data going back much further, so it doesn’t make sense to constrain it to the limitations of the simulation engines ( limited by the Compustat license ).

2-3 hours of analysis convinces me that some of the FRED time series that are not available to rules would likely be very helpful for market timing. Any update on making them available?


Agreed they would be useful.

P123 Team, More than 2 years ago you made some FRED data available to rules.
When can we expect the remaining FRED point-in-time data, like unemployment rate, etc. to become available for rules as well?

You said: “It’s on the to-do list but it has been pushed aside several times because simply adding it to the current sim engines would only simulate from 1999. However we have data going back much further, so it doesn’t make sense to constrain it to the limitations of the simulation engines ( limited by the Compustat license ).”

Does the new 15 year data limit affect this in anyway? Should not be a problem since one would probably only use it with ETFs.