nifty graphic from ofDollarsAndData.com

from OfDollarsAndData
I guess beware of expensive stocks...and try to buy cheap.

On how to characterize my blog...not to mention your blog

In a book review on the clandestine literature of pre-revolutionary France ("A Taste for the Taboo" by James Penrose, a review of Robert Darton's The Forbidden Best-Sellers of Pre-Revolutionary France; The New Criterion Feb 2018) I ran across this quote from Diderot about his Encyclopédie  which I found to be an apt description of blogs in general and mine in particular:

[it was a] "sink into which [were] haphazardly tossed an infinite variety of poorly digested, good, bad, disgusting, true, false, uncertain, but always incoherent bits."

No doubt I quote this not really to describe blogs as such but because I secretly desire to show my children -- in a desperate last minute attempt to role-model before they go off to college, a time point at which it might just be a tad late for such things -- that I read something (anything!) other than SSRN.com pdfs on geometric mean analysis, perfect withdrawal rates, and pension accounting.

Notes on my excellent spending adventure

1. The Problem Zone

If you've been following the blog you know I've been playing around with some spending simulation and math lately.  This has been in order to get myself more familiar with the more advanced spending PV calculations one might do in a plausibly academically-grounded and robust household balance sheet (HHBS). The balance sheet is a tool I lean on pretty heavily for managing my retirement in what looks like an increasingly post-human-capital world.  In addition, given the several types of simulation tools I've built over the last year or two, along with the recent HHBS work, I wanted to give myself a much better intuitive sense of some of the differences between the concepts of feasibility and sustainability.  In this self-directed learning process I have built a couple rudimentary spending simulators that kick out a giant piles of spending NPVs so that I can do statistics on them such as the mean (often used when defining stochastic present value [spv]) or percentiles or rank since the distribution can take on decidedly non-normal shapes in scenarios with high variance and summary stats other than the mean can provide interesting insights in those particular situations.

Since I don't have a serious background in statistics, probability or calculus and my last formal math class was in about 1977 (if you ignore grad school) I've been casting about on and off and here and there for a year or more for help on notation and concepts that are probably a step or two above what I can extract from Google. It's amazing how little access I have these days to the level of help I need and how often I have to revert back to a search engine. But I have had some successes of late. In addition to the feedback I've received from the small wins, I have also written notes to myself along the way or drafted emails using some of the material below in order to get prepared to communicate my questions to others (e.g., a pension risk director at PwC or a retired mathematician).  This post is a more organized version of the sum of the various self-notes and emails.  All of this is likely of interest to no one but myself but the process of integrating and posting it helps me codify this stuff in my own head.  If someone else finds it useful, so much the better.  Some of this is redundant with a past post or two.

A day in the life of Americans

More "theory reference points" (Yaari) for my custom spend calc and stochastic present values

As my "hobby" now seems to be writing a retirement finance theory blog, a topic on which I have no formal training and a field in which I have no deep experience, I find myself reading a lot of retirement-finance related technical or economic research papers. This, in other words, and if we were to look closely and be honest about it, is a "hobby" that my kids and my girlfriend find appalling. In fact I suspect that I probably could find better uses for my time like taking a tennis lesson or an art class or cleaning the house (please, they beg me) or maybe even, heaven forfend,  looking for a post-retirement job (always open to offers btw).... But as long as I still have "my thing" going I have to mention that I happened to notice that in the last 10 or so papers that I've read, maybe even the last 20, 100% of them point back to Yaari (Uncertain Lifetime, Life Insurance, and the Theory of the Consumer. 1965) or at least they do in the bibliography. I felt obligated to take a look especially since the papers I'd read kept referring to Yaari's math on consumption paths and subjective discount rates. Those are two things I've been thinking a little bit about lately.

Option value of not annuitizing? - another amateur hack attempt

One of the main themes of RH has been that an amateur, armed with nothing other than curiosity and the willingness to embrace some basic math and technology, can gain at least some minimal insight into their otherwise fairly complex retirement problems without the intermediation of "experts" such as potentially conflicted and maybe under-trained financial practitioners or remote, opaque, ungrounded, abstract, and non-skin-in-the-game academics.  I've used that approach often here to get some fairly certain understanding of some core real-world processes that have an impact on retirement.  But maybe here, today, with this post, I have extended myself a little too far out over my skis.

Probability distribution of a net wealth process over 20 periods

A RH recurring theme: I wanna see what "it" looks like.

Today's look-see: probability distribution of a net wealth process over N periods

In this case I am doing a very abstracted or stylized process where wealth is in wealth units (e.g., $1M/$40k = 25 units, which in fact are the units we start with here) and the spend is a constant spend of 1 unit  so

W_t =  mW_t-1 - 1.  

Return m in this case was an arbitrary 2% real net return with a 10% standard deviation (I call two percent "arbitrary" but I just read a paper where some semi-reputable ret-fin guy made a case that an expectation greater than 2% real over forthcoming years starting in 2018 might be more unreasonable than not. The 10% std dev I just made up. We'll go with these assumptions for now because it makes an interesting illustration[1]).  Start age is irrelevant because we are looking at the net wealth process which is independent of longevity.  Given the process for W_t ,  I simulated it over 20 periods (actually more than that but 20 makes the charting easier) 4000 times. Off of that sim I created an empirical density from the frequency distribution of wealth in any period t. The chart looks, without comment, like this:

Weekend Links - 2/16/18

Source: CNBC
Too close to home.  Kids from Parkland go to my kid's school.  Children of my kid's teachers go to Parkland. Parkland is a few miles from here.  Enough seems like enough, doesn't it?


QUOTE OF THE DAY

I’d like to think generosity and manners are a signal and cause of success.  L2   

RETIREMENT FINANCE AND PLANNING

Headed for the Poorhouse: How to Ensure Seniors Don't Run Out of Cash Before They Run Out of Time, MacDonald, Ryerson/NIA [Canada]

I propose a government-led solution: Canada’s Living Income For the Elderly (LIFE). As an integrated component of the Canadian retirement income system, LIFE would effectively enable retiring Canadians to pool their financial savings to better protect those who live to age 85 and beyond. LIFE would give Canadian seniors the affordable, secure retirement income they want, when they need it, without shifting the cost and risk burden to the rest of Canadians. [If I were Canadian, and I do have Canadian roots btw, I'd be wary here…] 

Report proposes new retirement income program[Canada, see last link], investmentexecutive.com

The C.D. Howe report proposes a government-led Living Income For the Elderly (LIFE) to overcome these obstacles. The program would involve creating “a national, restrictive, non-cashable, advanced-life deferred annuity with non-guaranteed (but conservatively targeted) payment amounts, with potential end-of-year income ‘bonuses’,” the report says. [as above, I am withholding judgment but I'd be wary].  

Financial Economics Principles Applied to Public Pension Plans [2016], ssrn.com

Working from basic principles of economics, financial economics, and public finance, we develop implications for the financial management of public pension plans. We address the measurement of plan liabilities and cost, funding, investment of plan assets, financial reporting, benefit design and risk sharing. Our analysis seeks to maximize efficiency and preserve intergenerational equity. We conclude that full funding based on default-free discount rates is efficient and fair across generations. Investing so as to hedge accrued liabilities facilitates the maintenance of full funding across time, minimizes risk-adjusted costs, and avoids potentially costly and/or futile risk taking. 

An early-retirement-stochastic-PV game

This post takes off from a previous post on stochastic present value(SPV) where I was trying to look at things from the point of view of an early retiree and think about what my future SPVs might look like from today's standpoint (I don't know, maybe I am budgeting for future balance sheets or mindlessly fearful or just messing around with numbers.  I doubt there is a practical usefulness here).  In other words, if I projected/inflated my current spending x% to next year (3% in this example) and then ran a SPV calc as if I was starting from next year's stand point with entirely new longevity assumptions, what would the SPV look like in next year's dollars...and then I inflated today's spending two years to the year after next year with that "platform's" new longevity expectation and so forth.  Is this legit? No idea.

Simulated stochastic present value of spending meets an efficient frontier

This post falls in into that category of "let's check out what it looks like before we have any idea what it means."  That "let's check out..." thing also appears to be a recurring theme here.

In this case I wanted to take the stochastic present value idea -- i.e., take a deterministic spending present value calc and animate it with random variables for the numerator (spending, inflation), the denominator (discount rates representing the return generating forces at play), and duration in terms of longevity and use it to create a distribution of spending NPVs for use with a household balance sheet -- and then see what happens if we connect it to an efficient frontier of asset allocations[3] for a given set of arbitrary assumptions.  That is a mouthful.  Let's take a look.

On the affinities between retirement tools and optical systems

Somewhere on this blog I have a post about optical systems.  I was once a telescope builder (four of them if I am counting right, here was one of them).  One of the things that most people misunderstand when they look into the eyepiece of a telescope is that what they are looking at is not actually a star even though they think it is. That sounds like it is at odds with common sense but here is how it works.  Because of the geometry and distance involved, a star is too far away to actually be able "see" the disc of the star like one can see the disc of a planet (I suppose it is possible if the telescope is big (wide) enough or the star close enough and I should double check the current science to see if someone somewhere has done it).  For small telescopes what one sees is actually the aftereffect of a wave-front of light (remembering here that light has "dual nature" as both particle and wave) interacting with the aperture of the instrument.  The aperture (the opening at the front of the telescope) is like an opening in a sea wall.  The wave front is like, well,  like a wave front.  The wave hits the opening and bends around the corners and creates wavelets (diffracts). At some point near the beach if one were to "look," the combinations of the various waves creates high spots and low spots (interference).  This is also what happens to light (or, rather, it is a pretty flat-ish wave front at that point) from a star.  It hits the (usually round) aperture* of the telescope and diffracts and when the image is examined at the eyepiece the combination of diffraction and interference means that the waves are combined to a really high point at the center (i.e., a bright point, you can see it) then there is a dark ring around the center where the interference cancels things out and another fainter light ring around that and another dark ring and another faint light ring and so on.  It gets geometrically fainter away from the center so what one sees is mostly the center diffraction/interference peak (called an "Airy disk" for the guy that wrote the first theoretical treatment of the phenomenon) or a combination of the first couple rings, not a star.  What one "sees" is entirely an artifact of the interaction between an aperture and a wave front of light. In fact, counter-intuitively, the smaller the aperture, the larger the "disk." One is not really "seeing" a star at all. It's an an illusion of sorts.

I propose that something vaguely similar happens when using retirement tools to examine retirement risk.  In this metaphor the instrument or aperture might be software/math while the wave front is data and assumptions (or maybe the assumptions are part of the instrument? I don't know. I guess I haven't thought about it carefully enough).  The image that is examined would be the output of the system: fail rates, "optimal" asset allocation recommendations,  glide paths, and so forth.  The point here is that the results created by the interaction of instrument (software) and wave front (data) creates an illusion that looks like retirement or the future but is nothing but an artifact of the interaction between the two components.  It may be useful but it is not "real" especially if one is very far away from the object of interest. On the other hand the closer one gets to the endgame the easier it is to see and judge the risk  Maybe some of this metaphor breaks down somewhere in there because the wave front of light is real and live where the data used in planning is usually past and dead but I'll have to work that out.


----------------------
* Think in terms of Newtonian telescopes here where the light is un-intermediated at the aperature and no refraction through glass is involved at or near the aperture to complicate the point of the post. In other words, light is just going through a round hole.

The theory behind my custom spend calc?

The answer to the title: yeah, this is probably it...if I understood it better.  I was thinking about this because I was recently perusing a Society of Actuaries research report Value of Longevity Pooling, Feb 2018 -- and at this point it would be reasonable to think that someone somewhere might be wondering about what might be missing in my personal life that I find myself down that particular rabbit hole -- and I found some good material from Moshe Milevsky (always a good source for complex and relevant things in lifecycle finance) on "The Value of Mortality Pooling: a.k.a. Approximating Annuity Equivalent Wealth (AEW)." Down at the bottom of this particular hole I found this:

Perfect Withdrawal Rates and Asset Allocation

In a past post or two that I am too lazy to find for the links I recklessly claimed that for a broad range of retiree asset allocations in the center mass of risk the exact allocation doesn't matter a ton or at least doesn't matter much when compared to bigger fish that can be fried like spending.  At least I made the claim in good faith based on: a) a bunch of reading of and interpretation of other's work, and b) my own sorta-semi-legit attempt to come up with an optimal allocation framework for age and wealth levels using backward induction and stochastic dynamic programming (borrowed idea and implementation framework; my code) and then sticking the results of the optimization into a simulation framework to see what works best.  What worked best, when non-rigorously optimized for both fail rate and fail magnitude in years, was a range of allocations (two simple asset classes were all I could pull off at the time) between 40 or 50% equities to 70-80% equities. For the whole range from zero to 100: less than 40 had dramatically bad news, 40-70 was more or less all the same and good enough, and greater than 70 or 80 was ever so slightly less optimal than 40-70 but not nearly as bad as less than 40 and in some cases, depending on age and level of wealth, was perfectly ok all the way to 100%. You'll have to take my word for now that this is, in fact, what I had concluded at the time.

Deconstructing one non-normal SPY distribution into two normals -- just for fun -- and then putting it back together again

I thought I'd take the distribution mix concept I've been working on out for a test drive.  I realize this is kindergarten level work for a quant but it's a new tool to me and I'm trying to get the hang of it.  To do that I took the full yahoo history of SPY returns and used expectation maximization to deconstruct it into two normal distributions just to see what came up and what it looked like.  I then used the two normal distributions to turn around and then reconstruct a fake fat tailed skewed distribution like SPY. The idea is not necessarily that I want to model SPY -- but I might -- it's that that means I could model a pretty broad range of other things that do and do not exist.  Which might be fun.

Running the algorithm (not my excel or R model this is an R function called normalmixEM()) came up with SPY being ~83% a normal distribution [EM dist1] that has mu=.0166 and sd=.0333 and ~17% a normal distribution [EM dist2] that has mu=-.0302 and sd=.0525. The original SPY was mu=.0085 and sd=.0414   These are monthly series, by the way.  When you reconstruct it you get a gaussian mix with mu=.0072 and sd=.0446. I didn't measure the other moments yet because there is something wanky with my understanding of the two competing R functions I was using to do that and I didn't want to get it wrong.  On the other hand visually it works out nicely....like this:

  - black is the original SPY density for monthly returns
  - blue is the normally distributed "EM dist 1" (high) - random return generation
  - red is the normally distributed "EM dist 2" (low) - random return generation
  - black dotted is the artificially/mathematically reconstructed non-normal Gaussian mix

Works pretty well (smallish data series so not perfect).   I thought that was pretty slick.  In fact after I posted this I was thinking about it a little more.  I am, for better or worse, an amateur or perhaps a tourist visiting the land of retirement finance and probability theory. I have my camera and my ugly tourist shorts but that's about it; I don't speak the language and I don't live there. So, for me, while it is one thing to know some basic stats like the various moments of a distribution or how to generate a CDF or how to integrate a PDF, its another thing altogether to look at a distribution and see multiple other distributions hidden inside trying to get out.  I'll probably never look at a data distribution in quite the same way again. I'll call this whole exercise a "net add" to my trip.

Weekend Links - 1/8/18

QUOTE OF THE DAY

The benchmark is “my family and I are alive, safe and fed”. The rest is luxury. Daniel Egan  

RETIREMENT FINANCE AND PLANNING

What You Can Learn from Fighter Pilots About Making Fast and Accurate Decisions

Farnam Street

If you are able to be nimble, able to assess the ever-changing environment and adapt quickly, you'll always carry the advantage over your opponent. 

Unraveling Retirement Strategies: Floor-and-Upside, Cotton

In between these extremes, the decision can be more difficult. The best I can recommend is that you imagine that you are 85 and your upside portfolio balance just went to zero, a victim of sequence of returns risk. What is the least amount of income you could have remaining that would not make your life an economic misery? This is the floor level you wish to have. 

What's a Floor? Dirk Cotton

I begin with the goal of a floor portfolio that provides near-certain safety-net income and I try to fill it with assets that in combination mitigate inflation risk, capital market risk, and longevity risk to guarantee that I can survive improbable but worst-case outcomes. Because floors are expensive, I build mine as low as I think I could tolerate and then I structure the rest of my retirement plan to minimize my chances of rolling off the bed. 

Second Childhood  Jonathan Clements

As a gut check, I use the strategy I recommend to others: Occasionally, I will take my portfolio and assume the stock portion loses 35%, which is the typical decline during a bear market. I’ll then look at the resulting hit to my overall portfolio’s value and ask myself, “Would you be okay with that?” [yes, this is, in fact, the correct question to ask oneself] 

Working with Retirement Income Buckets – Webinar Q&A, Pfau

Our recent webinar on the topic of retirement income buckets, or time segmentation, was a big hit with many questions coming in. In the past I’ve tried to reply personally to as many of the questions as I could. But with travel this week, I became overwhelmed. Also, if one person asked a question, probably many people had the same question in mind. So I thought I’d instead provide a list of questions and answers for everyone to be able to read. Let’s get to it… 

The slowing pace of life expectancy gains since 1950, Tyler Cowen

in many United States counties, life expectancy is moving backwards these days. 

Original link here https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-018-5058-9

Hyperbolic Discounting Can Be Good for Your Health, Strulik & Trimborn

We set up a life-cycle model of human aging and longevity in which individuals discount the future hyperbolically and make time-consistent decisions. This allows us to disentangle the role of discounting from the time consistency issue.  We show that hyperbolically discounting individuals, under a reasonable normalization, invest more in their health than they would if they had a constant rate of time preference. Using a calibrated life-cycle model of human aging, we predict that the average U.S. American lives about 4 years longer with hyperbolic discounting than he would if he had applied a constant discount rate. The reason is that, under hyperbolic discounting, experiences in old age receive a relatively high weight in life time utility. In an extension we show that the introduction of health-dependent survival probability motivates an increasing discount rate for the elderly and, in the aggregate, a u-shaped pattern of the discount rate with respect to age. [the sharp-eyed will recall my suggestion that hyperbolic discounting might be a useful tweak to retirement PV analysis. I can't follow their math but I get the general point] 

I solved my recent EM Gaussian mix problem, but...

Ok, so I solved my recent EM (expectation maximization) and Gaussian mix quantitative finance problem...but now I am irritated. And it wasn't easy for me.  But let's get this out of the way first.  This was my purpose:

1. I just wanted to learn something new and complex for the f*** of it because that is what I do. We all need our strange hobbies, and

2. Being able to describe non-normal distributions mathematically enhances my ability to do financial modeling in life-cycle simulation and some other things I do by quite a bit.

But here is my irritation.  I am 60 and my brain and eyes are decaying daily.  Searching and reading 1000 false leads online, which might have been fun and easy at 25 is no longer fun or easy.  I can't read screens for long periods of time and even if I do I have kids that are hungry or need laundry done or forgot to tell me they need poster-board for a project tomorrow or need a ride to a friend's house or, god forbid, are sick. Then, even if I have their cooperation and I have good eyes and time, I do not have access to work colleagues or professors or computer scientists or whatever else. This is what I have: I have my cats; I have the checkout people at the grocery store; I have my daughter at Stanford; I have reader David C in NY; and that is about it.  When I want to figure this stuff out I have to beg strangers in the world by email (usually no response) or send requests to yahoo answers or Ask Dr. Math (totally hit and miss) or, most often, figure it out on my own but with the constraints described above. 

Question for any reader on Expectation Maximization and Gaussian Mixtures

I've tried this kind of question before with exactly zero success.  I mean I have had help with ideas and content and paper recommendations and so forth but even that was only really maybe two readers with whom I have interacted in any substantive way (good luck with the forthcoming baby btw, DC), three if we include my sister and four if we include my girlfriend rolling her eyes at any of this. Zero help with the exact math though. 

I am currently working on what I'll call "gaussian mixtures for dummies" (success) but now also trying to decode a simple expectation maximization problem...but I am at a wall (fail). This is beyond my aging brain. Anyone know a simple way of helping me figure out how to back into the mix equations for my return series?  The stakes are pretty low, this is personal learning not professional.  I'm looking to do it in excel if I can, R if I have to and I am trying to avoid YouTube or google.  My eyes weary at the effort of searching. Even a pointer to a paper that could walk me through it like a 5 year old would help.  enelisvia at g mail. 

Why, exactly, are people losing their minds over this correction?

Remind me again why there is all the angst about this correction? This is the first one where I have not even looked up from my desk.  I remember, of course, standing agape, slackjawed, staring at something like what unfolded on the television screens in 1987, and I also realize that "On a percentage basis, the change in the VIX [this correction] was the single greatest increase over the last eleven years." But still, really, cmon where is the perspective.  Here, look at a monthly chart of the SPY over the last 15 years...

1. We haven't even returned to the 12 month moving average,

2. We have not broken the bigger trend,

3. We blew out of and exhausted the top of a channel (if you believe in channels) but haven't really even re-entered the core channel.  We did not lose money, by the way.  We had money last month that we probably shouldn't have had and had not really earned. We should never have even registered the idea in our head,

4. If you had told me in March of 2009 -- and I can still almost literally feel the nausea of the final free-fall flush down that quarter because I was in the middle of a move, a divorce and a retirement -- that we would be where we are EVEN AFTER THIS CORRECTION, I would have kissed you on the lips. 8% compound from the bottom (if I got that right and that would have been a lovely enough return all by itself...and closer to long term expectations than 16%) would have been something like 137 not the 266 I just saw. We should all be cheering rather than weeping and gnashing our teeth,

5. The american economy, as a trading partner used to tell me, is not going out of business and even if it were, you'd have bigger issues than emotionally processing a minor correction -- like where to get water, canned food, and ammo.   We are a tiny bit far away from that type of thing. 

Call me when we break the trend.

qqplot of my strategy

I hadn't done a qqplot on my strategy before but since I'm working my way through a quant-fin self-study effort and I hit the qq page in the book I have I thought I'd kick one out to see what it looks like for something I work on daily.

The red line is the monthly returns of my systematic alt and the blue is a random normal with similar mean and vol.  (let's call it 7% mean 4.1% sd if annualized).

So a little misshapen distribution.  I was hoping it would have been even odder but there it is.  The best I can guess on interpretation is that the tails first thin and then fatten with slightly fewer extreme moves at the ends.  The qualitative version might be that it looks like I give up some probability of some of the upside for a little bit but not quite enough diminution of the down compared to my goals.  But I really have no idea. Any quant geeks can help me out here... 

Hindsight 9: Three Independent Things

I woke up to retfinophobia -- defined as the fear of or unease about one's financial future in retirement...although now that I made that word up I suppose it could also mean fear of retirement finance content...like this blog -- in 2010 when I started to ask myself the deceptively simple but devilishly difficult (impossible) to answer question: "yikes! am I going to run out of money?"  The attempt to answer that question, which is destined to fail since it attempts to predict the future, involves some complex thinking and math while it also leans on a melange of many things all stewed together to come up with some assessment of "how much risk do I have" or "how long will the money last" or "how much do I need now."  This is where I started in 2010, a start that was initially inter-mediated by others but then slowly was taken onto my own shoulders as I started to learn more over time. This was also, by the way, the genesis for this blog.  I realize now, however, that some of that early time might have been better spent understanding the individual components of the problem, each of which are independent of the others, rather than confusing myself with the opaque and naively-understood blend of all of them. The blend would have made a lot more sense -- and is quite useful once known -- if I had learned the basics of the independent processes first.

Backward looking mean-var for my systematic alt risk strategy - update Jan '18

Note that this looks backward and not forward and I'm never sure if I have everything exactly right in the math. In particular I am not very confident in correctly projecting from one horizon to another.  Also the code is aging and reading it is now hard if not impossible so I hope my work in building it back whenever does not need any more scrutiny at this point. On the other hand everything is playing by the same rules here so maybe it's ok.

The green dot is the strategy I run myself that can be characterized as a systematic alt-risk strategy that uses mostly fixed income category ETFs with a rules-based trend following overlay. There is also some credit risk in there and short options (vol risk prem) with an occasional macro-discretionary thing. This is updated thru Jan 2018 and begins in 2013 which is where I have the most consistently reliable data for my stuff though the strategy itself originally dates from 2010.

The 5-asset portfolio, for better or worse, is SPY AGG IYR GLD and EFA with allocations selected via a sampling technique.  Returns are linear annualized averages which I'd have to look again to see how I did it.  I convinced myself at one point it was legit but now I'm not so sure. The portfolio std dev is projected using a morningstar method that expands the usual rule of thumb a bit.  The hedge fund index is Barclays only because that is all I can get for free.  The allocation benchmark is an allocation etf (AOM, total return) which is approximately 40/60 if I recall correctly...also free.  the 40/60 dot is that allocation of SPY/AGG.  Advisor is a strategy I have managed for me but is net of fees and witholding, etc -- also time weighted -- so a little bit of apples and oranges.

Lesson? The green dot is only maybe 15% of net assets so the impact to me is not huge but on the margin I still believe in the ability of a systematic alt risk strategy to add some value either on its own or as a diversifier that adds a little incremental efficiency to a bigger portfolio though I have not directly measured the covariance of the green dot vs other stuff to be more sure of exactly how it works.  The value-add (or destruction) also depends on ones portfolio and risk policies.  The other lesson is that: a) it is harder than it looks to stay fully invested in the strategy (actually I prefer being a little bit levered), and b) it's more work than I want in retirement even though I have automated a ton of the work.  I think I'm looking to close this down or outsource sometime in 2018 or 2019 to save some time and maybe beat strategy decay if that were to be something to anticipate in my future. 

Some fun software tricks in image processing

This is a little off topic and I'm thinking that the technology has moved on and up from where it was 14 years ago when I did this kind of thing but I thought this (below) was a pretty neat trick at the time.  "At the time" was the time when I thought I was going to go big in astro-photography, somewhere around 2003.  The combination of high quality optics, digital cameras (yes they were new technology at one point), and image processing software was opening up opportunities for amateurs that would not have been very easy, or even possible (?), a few years earlier. 

In the old days one needed giant telescopes that gathered a ton of light, near zero light pollution, and an extremely good mount so that one could take very long low-light exposures of very faint objects with no tracking error and no shaking.  The new technology, in simple terms, meant it was now (at the time) possible to do something similar by taking many short digital exposures, stacking them electronically, collapsing them together and then subtracting out the noise.  That approach reduces mount tracking errors, the impact of light pollution, and also means smaller less expensive and easier-to-handle telescopes can be in play. 

Rudimentary spreadsheet sim vs. imperfect estimator for N-period geometric mean and sd

...or a re-title might be "why did I feel like I had to simulate other than to see how it works?"

I have a few spreadsheets and R-scripts that model the behavior of geometric returns over finite time horizons.  I did this mostly to educate myself and because I think mid-tier time horizons like 15 or 20 years can be of practical interest to retirees where that kind of horizon might be less interesting to a younger accumulator or an institution.  I know from what I've read (Michaud, Mindlin, etc) that there are estimators out there that can do the same thing as the sim just fine, but I just have not used them much.  Today I was curious how big a difference (not measured exactly, just visually compared) there would be between a simple spreadsheet sim that uses random normally distributed returns as part of the return generation process and the estimator offered by Michaud in A Practical Framework for Portfolio Choice [2003].