No. of Recommendations: 17
could you perhaps explain or maybe point me to a previous post on how you arrive to your smoothed value figures?Really short answer: four year average of 1.5 times book per share.
Longer answer:
As above, but with a few tweaks, including:
More recent data is more important than older data, so I use a "weighted moving average".
Weight of 1 on the oldest data point, weight of 2 on the second-oldest, weight of 3 on the third oldest, and so on up to the most recent figure.
I used 16 quarterly figures instead of 4 annual figures, with weights 1 through 16.
For each quarter I used the average of (1.5 times book per share) and (my two-and-a-half-column valuation method).
It makes surprisingly little difference compared to simply using book per share, but my valuation method is a bit less volatile.
I am of the opinion that drops in book value are always transient and do not represent drops in the true value of a share, just a limitation in the precision of book as a value proxy.
However, I'm willing to admit that sometimes just before a bear market, book per share might be a little bit unsustainably high when stock valuations are bubbly.
Remember the price of Coca Cola in 1999, for example.
So I picked a figure of 96% of peak-to-date book-per-share as the "floor" of my estimate of value.
For each quarter, I used 1.5 times the higher of (current book per share) and (96%-of-peak figure) so there are dips in bear markets, but never big ones.
That figure is then fed into the 16 quarter weighted moving average described above.
After you do this, and especially after my tweaks, the smoothed value is always rising.
It just slows down during bear markets - a lower rate of growth of observable value.
The year on year increase in smoothed value got down to inflation + 3.3% for one quarter in the credit crunch. (recent peak inflation+12.9% 2018-Q3)
Without the smoothing, the year-on-year figure would have hit inflation -4% three times.
The smoothing makes the slowest periods seem slow, not negative.
Here are some old posts on the idea of using a smoothed value proxy as an input to a safe withdrawal rate calculation
Original post
http://www.datahelper.com/mi/search.phtml?nofool=y...A follow up 5.5 years later (2.25 years ago)
http://www.datahelper.com/mi/search.phtml?nofool=y...Jim
