I like the overall logic of the CAPE method for smoothing variable spending in response to market conditions, but I'm unsure of how to use it in a practical way to plan variable spending for the near term (e.g., next year). I understand that the historical sequence populates the tabular projections with variable spending based on some specific historical sequence, but is there a way to simply plug in the current CAPE/CAEY, along with 'a' and 'b' values to project a variable spending level for the coming year? One alternative is to use ERN's spreadsheet, but I'd have to duplicate a lot of the information (in far less detail) that is already captured and calculated by Pralana already. Am I thinking about this wrong?
Thanks for any help and insight.
@jdphog I'm sorry but, no, there is not a way to do that. The primary focus of Pralana tools is to make long-term projections. A hypothetical question would be: if you plugged in the current CAPE/CAEY for this year (to project variable spending for next year), what would the tool use for CAPE/CAEY for future years in making its projections? This is why the CAPE method is only used in doing historical sequence analysis where an entire sequence of returns and CAPE/CAEY values are available. So, here's my best thought on how to achieve your goal:
The CAPE and "a" and "b" values are used like this: withdrawal rate = a + b/CAPE for the current year. You could do this calculation by hand and then plug the resulting withdrawal rate into the Fixed Spending % strategy to get the variable spending amount for next year.
Stuart
@smatthews51 Thanks Stuart. I'm thrilled with the long term planning focus, and you're right that it's not hard to calculate the current year's variable spending level using the formula.
Just one follow up question/suggestion. Might it be useful to allow the user to enter a current year CAPE value followed by a specific historical sequence? For example, I would enter this year's CAPE to calculate variable spending for the coming year and then model what would happen if the 1966 sequence started the following year.
On a related note, what is your source for the historic CAPE values? I'm wondering about this in relation to ERN's Building a Better CAPE Ratio article.
Thanks,
Jay
@jdphog Regarding your suggestion of allowing you to enter the CAPE for the first year and then use some specified historical sequence for the subsequent years, I suspect the value for the first year would yield virtually no long-term difference from simply using a specified sequence for ALL of the years (which is how the Historical Sequence Analysis currently works.)
The source of the CAPE values in PRC's historical database is Robert Shiller's data.
Stuart
