I know I'm more oblivious than most so I am probably the last reader of finance to have come to this kind of epiphany. For several decades I was under the illusion that the "amount" or the number of things is what matters, which it does of course in some essential way, but really it all seems, in the end, to be a little bit more about time. I mean, yes, in grad school we learned all about managerial accounting and present value analysis and yield curves and stuff like that but any discussion of continuous-time finance or geometric returns or other time effects was more absent than not. MBA is not really where one learns that stuff of course -- or it didn't used to be in the 80s -- but it seems a little odd now in retrospect that I missed it.
I'll keep this brief and will position us in the context of retirement finance and this blog. In retirement finance blogs or papers or articles or practitioner marketing, the focus usually seems to be generally on stuff like: one's "number" or the "fail probability at 30 years" or the terminal wealth distribution or percentile or a bequest likelihood or amount i.e., the talk about "amounts." But, lo and behold, other things have now come into focus for me over the last 8 years (keeping in mind my potential obliviousness). There are probably a million things to discuss but here is what I am specifically thinking about off the top of my head:
- portfolio longevity denominated in years. This is the granddaddy topic and quite underappreciated. I started here when reading in 2012, Milevsky's "Seven Equations..." book. Paired with longevity it makes a great conversation
- spending, but really as a flow or a rate per unit of time. Rates are an interesting subject. The perseveration on the monolithic "4%" topic distracts us from the reality of the consumption rate topic at a meta level. Maybe I over-state this, idk, but once we get into spend we really really get into process and once we get into process we get into time. Alfred N Whitehead might be proud of me on this but who knows?
- we can calculate the fail "probability" at some horizon (30 years seems to be the thing but that's not really time. What it is is "one period" 0 to 30 which is not time or a process but a single unit), sure of course, but now also...let's do a magnitude in years. A fail at 102 with one breath left is benign. A fail at 69 with another 33 years left is kinda malignant. Most wise paper writers will and do point this out. The naïve jump over it
- discounted utility of lifetime consumption. This evaluative method is nothing more than a "measured consumption fail process" except now with a focus on consumption over time rather than wealth at a point, and with ambient income (a process) and with a convex penalty function that captures the magnitude of the time within the diminished consumption state. So, basically this is the same as the magnitude chat in the last bullet, just more robust and realistic imo...if one likes utility math that is
- geometric mean return over multiple periods or maybe alternatively log returns over single. I still marvel that this was not discussed in 2 years of grad school finance. This still seems like a big miss but guess that is not what MBAs learn
- conditional survival probabilities and hazard rates that necessarily align with age and time. If life itself is not a time-contingent process, I don't know what is. It's a pity that more quant retirement finance papers do not introduce this concept. Probably why I am a Milevsky "fan boy" heh, it is a major topic of concern as it should be
- continuous time finance which brings time to an infinitesimal, and probably necessary, point not all that unrelated to the point on geometric mean returns. I should be clear here that I know this topic almost not at all. It does not surprise me that an MBA program that churns out a s-ton of marketing majors and proto-treasurers does not dive into this topic. Maybe kind of glad they didn't but at 60 it got more interesting.
No comments:
Post a Comment