Announcement

Collapse
No announcement yet.

Juggling Trinity Study Data AOF Style

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #76










    You say AOF isn’t trying to predict the future, isn’t accurately describing the past, and is instead describing what might have been.   That is the definition of beyond useless to me.
    Click to expand…


    What do you think of the life cycle (nalebuff/ayers) approach to this? It kind of ties in with dynamic allocation/glidepaths. We all know that higher equities allocation actually gives one the best chance of the longest withdrawal period and highest ending or spending allowance. However, that comes with extreme SORR. You have the majority of your wealth in just the last decade or so.

    Supposedly, and the math seems to work, using the life cycle way they espouse and “diversifying across time”, which is a pretty eloquent solution or at least illusion, you have similar downside risks as to a static portfolio, yet the average portfolio and best case scenario results are significantly better than a fixed scenario. Nice, if you can take the 100%+ equity approach. Few can do 200% for considerably long periods of time, but I am trying to think of it as filling up my 80-90% equity bucket first, then filling up bonds right at the end.

    I will add the caveat I do not think this is a good strategy for your average retail mom/pop investor. Prudent account management and paying attention to your leverage and having a systematic plan makes sense as well. I dont think there is a single person that can truly take a whole portfolio massive draw down just sitting there, no matter if they understand its early/not much overall % of retirement wealth and they will be better long term. Humans dont work like that, I know I dont. So stops or some kind of less risk or risk off asset makes sense.
    Click to expand…


    I think people do this life cycle investing with real estate. They take on high debt early and gradually pay it off over 15-30 years. Unlike a loan on stocks there isn’t generally a risk of margin call, only default if you lose your job.

    In terms of life cycle, you are most able to absorb hits early in your life (before kids) or after they have grown up. Having high debt when you have kids is stressful. Maybe it’s better for your final net worth but may not be worth it.

    I can’t see why you can’t take more risk at the end of your life. If you have enough for retirement at age 55, what stops you from taking some of the excess and leveraging it if you want to swing for the fences ? In fact if you have excess for retirement at any age, you can leverage up what you can afford to lose.

    Hopefully though you’ll have more experience and not blow up. Taking on margin loan debt at a young age is a sure fire way to blow up. But maybe you’re better off blowing up at a younger age and learning from this than doing this at a later age.
    Click to expand...


    The RE aspect is absolutely true but people dont think of that as leverage of course. This is what people are saying when they say RE is the best path to wealth (those that survived), they are really saying leverage is the best path to wealth. RE has some nice legal and mark to market benefits. Leverage is also the best path to losing it all. Have to be careful.

    You dont have to be in debt to get a benefit from the lifecycle idea. A no debt version would be all stocks for the first decade or so, blended in the middle, and all bonds at the end or whatever worked out for your particular allocation.

    The point of leverage if you have well enough is actually true and the topic of I thought a Kitces article someone linked a while back. If you took reasonable leverage...who knows, 20%, you wouldnt put yourself in danger and would likely greatly increase your overall wealth. But as we age we get debt averse. Some leverage should not even approach levels that allow one to blow up. I think the article talked about a "prudent" level, but dont recall what they thought that might be. I mean if you were conservative but it made sense to you you could try a risk parity type and slightly lever up the bond side. The point would be to not do to much.

    Comment


    • #77
      Oh boy, I take a few days to try and get some ‘real’ work done and you guys are having an actuarial fiesta on Cinco de Mayo! I’m really not sure where to begin or even whether there is value in addressing individual points so I’ll respond thematically.



      What I Did?

      There are 4 types of simulations floating around the investment world.

      1. Monte Carlo sims. “What could happen”

      2. Back-testing with historical data. “What would have happened”



      3a. Re-sampling without replacement. This preserves the overall growth (CAGR) but changes the sequence of returns.

      3b. Re-sampling with replacement. This destroys any mean reversion in the markets.

      That’s the order of credibility in the investment world. I do monte carlo sims for clients all the time, it’s a well-established and credible way of setting investment strategy.

      Back-testing is generally frowned upon as it doesn’t tell you much about the future. If a bank or money manager approaches me with a back-tested strategy I don’t give it any credibility since they are adept at shaping the strategy to make it look good in the rear view mirror. However back-testing is loved by the PF community since it is simple to understand. I have done a couple of posts with this technique, for example this post on Short-Term Investing.

      3a was something I noodled over in a couple of posts. I liked it, some others did too, and others went into paroxysms of rage. See for example this post on buying a boat.

      3b was suggested as a thought experiment by the Nobel prize winning economist  Paul Samuelson and seemed like a fun thing to model and was the subject of the Rockstar post.

       

      Why I Did It?

      My main interest here was to see how investment strategies might change if we drop any assumption on mean reversion. Since this is such visceral belief by individual investors I thought reproducing the Trinity Study through this lens might be provocative and insightful. I think it was both provocative and insightful – mission accomplished.

      However the insight was surprisingly banal. My conclusion was that a ~4% SWR was probably fine around 95% of the time.

       

      What Don’t I Do?

      I don’t do any monte carlo sims on my blog. It’s not as simple as choosing a probability distribution and off you go. There are a number of fundamental assumptions that need capturing like correlations, mean reversion of interest rates, reversion of equity PE levels etc. And how are these set? Mainly by looking at the past. So actually #1 and #2 above are really not that different.

      It also means that all monte carlo sims over the last 10 years have been misleading, since they have built in mean reversion of interest rates, when actually they stubbornly stayed historically very low.

      I’ve done my fair share of peer reviewed research in the mainstream; journal articles on mathematics and white papers from large global investment consultants, but my blogging is supposed to be fun. Simply reproducing the financial mainstream goes against my blogging ethos. Which brings me to…

       

      Blogging Ethos

      I think some of you have been badly misled around what you can expect from my writing. WCI readers have been spoiled. You have a comprehensive and mostly accurate guide to all aspects of personal finance advice at your fingertips. It’s presented in a pragmatic, understandable way that is action orientated with an emphasis on practical guidance. e.g. here are simple steps to do a Roth conversion. That is a fantastic thing, especially when combined with a helpful forum.

      Think of it like… WCI is a roll-on/roll-off ferry that quickly and efficiently steams from point A to point B depositing large numbers of passengers and vehicles at the same destination. The AoF blog is a rowboat with an erratic schedule that might not even get to point B, and is likely to get diverted up meandering and interesting tributaries.

      If you got frustrated because you thought you were on the ferry, but found yourself on the rowboat then I can sympathize (but not apologize).

      Most of my posts will not contain actionable advice, but some might. Some of my posts will concern mainstream issues, but some will be offbeat. There is no blog ‘mission statement’.

       

      Technical Point

      I can’t resist making a technical point:






      When bond prices go down, the yields go up. When stock prices go up (assuming the fortunes of the company haven’t changed), the dividend yield goes down.
      Click to expand...


      These two statements are mathematically correct. The first is bond math - bond prices inversely move to yields. The second is just, um, normal math. Div_Yld = D / P. In contrast my article daringly broke a “rule” that is not mathematically enforced – the mean reversion of stocks.

      So you have simply presented two tautologies, and as far as a contribution to the debate in this thread -  I learned nothing from it. It was worthless to me.

       

      Comment


      • #78
        Hooray.

        Thanks for posting!

        Alot of your analysis is prompting me to sharpen my stat + programming skills so its a nice "side effect".

        Comment


        • #79


          The AoF blog is a rowboat with an erratic schedule that might not even get to point B, and is likely to get diverted up meandering and interesting tributaries.
          Click to expand...


          My thoughts...

          1.  Not All Who Wander are Lost.  Search WCI for my posts on the topic.  I like to wander.

          2.  I find AOF's posts fun and thought provoking.  He challenges norms of how I have seen finance presented elsewhere.  His writing makes me laugh out loud.  He combines finance with Avenger's references...awesome!  Everything I read in finance doesn't have to be actionable.

          3.  He gets that math is not a crystal ball into the future.  Donnie's, Zaphod's, and Complete Newbie's back and forth has reinforced to me that math is not a crystal ball.

          4.  Worth vs. value.  Value is subjective, and I find it in his articles.

           

          Comment


          • #80




            Hooray.

            Thanks for posting!

            Alot of your analysis is prompting me to sharpen my stat + programming skills so its a nice “side effect”.
            Click to expand...


            One of the things that just doesnt seem to go away for me is my wish that I had went further in maths, and kept up what little I did know. So important. I wish I knew something about computers as well, but maths, so much good stuff there.

            Comment


            • #81







              Hooray.

              Thanks for posting!

              Alot of your analysis is prompting me to sharpen my stat + programming skills so its a nice “side effect”.
              Click to expand…


              One of the things that just doesnt seem to go away for me is my wish that I had went further in maths, and kept up what little I did know. So important. I wish I knew something about computers as well, but maths, so much good stuff there.
              Click to expand...


              IDK about needing more advanced math for personal finance.  I'll stick with Ben Graham:

              "Whenever calculus is brought in, or higher algebra, you could take it as a warning signal that the operator was trying to substitute theory for experience, and usually also to give to speculation the deceptive guise of investment."

              Comment


              • #82







                Hooray.

                Thanks for posting!

                Alot of your analysis is prompting me to sharpen my stat + programming skills so its a nice “side effect”.
                Click to expand…


                One of the things that just doesnt seem to go away for me is my wish that I had went further in maths, and kept up what little I did know. So important. I wish I knew something about computers as well, but maths, so much good stuff there.
                Click to expand...












                Hooray.

                Thanks for posting!

                Alot of your analysis is prompting me to sharpen my stat + programming skills so its a nice “side effect”.
                Click to expand…


                One of the things that just doesnt seem to go away for me is my wish that I had went further in maths, and kept up what little I did know. So important. I wish I knew something about computers as well, but maths, so much good stuff there.
                Click to expand…


                IDK about needing more advanced math for personal finance.  I’ll stick with Ben Graham:

                “Whenever calculus is brought in, or higher algebra, you could take it as a warning signal that the operator was trying to substitute theory for experience, and usually also to give to speculation the deceptive guise of investment.”
                Click to expand...


                Except its not advance math but basic stats that we all glossed over. Remember that Biostats class in medical school? Yea me neither.

                However, it is essential. I know not everyone is into research, but I think basic stat knowledge should be requisite for ALL physicians. I am not into research, but like Zaphod, I wish I payed more attention in Undergrad.

                @Zaphod - nothing is lost. If I can do 1/2 an hour of brush up/coding a week with a 6 month old clinging to me + running side hustles, I think you definitely can! (you have the temprement to do it to boot I think...plus hey open invitation to post on a collaborative blog still stands!) But yea...do wish for more time or go back in time.

                Comment


                • #83










                  Hooray.

                  Thanks for posting!

                  Alot of your analysis is prompting me to sharpen my stat + programming skills so its a nice “side effect”.
                  Click to expand…


                  One of the things that just doesnt seem to go away for me is my wish that I had went further in maths, and kept up what little I did know. So important. I wish I knew something about computers as well, but maths, so much good stuff there.
                  Click to expand…












                  Hooray.

                  Thanks for posting!

                  Alot of your analysis is prompting me to sharpen my stat + programming skills so its a nice “side effect”.
                  Click to expand…


                  One of the things that just doesnt seem to go away for me is my wish that I had went further in maths, and kept up what little I did know. So important. I wish I knew something about computers as well, but maths, so much good stuff there.
                  Click to expand…


                  IDK about needing more advanced math for personal finance.  I’ll stick with Ben Graham:

                  “Whenever calculus is brought in, or higher algebra, you could take it as a warning signal that the operator was trying to substitute theory for experience, and usually also to give to speculation the deceptive guise of investment.”
                  Click to expand…


                  Except its not advance math but basic stats that we all glossed over. Remember that Biostats class in medical school? Yea me neither.

                  However, it is essential. I know not everyone is into research, but I think basic stat knowledge should be requisite for ALL physicians. I am not into research, but like Zaphod, I wish I payed more attention in Undergrad.

                  @zaphod – nothing is lost. If I can do 1/2 an hour of brush up/coding a week with a 6 month old clinging to me + running side hustles, I think you definitely can! But yea…do wish for more time or go back in time.
                  Click to expand...


                  I absolutely loved math in undergrad, but only took basics, alg, trig, calculus. Would have like a probability, real stats (not the med version), and some basic finance. Not necessarily anything crazy or advanced, just applicable. There is so much applicability to probability, stats, and some of the broader principles of say game theory.

                  Comment


                  • #84





                    The AoF blog is a rowboat with an erratic schedule that might not even get to point B, and is likely to get diverted up meandering and interesting tributaries. 
                    Click to expand…


                    My thoughts…

                    1.  Not All Who Wander are Lost.  Search WCI for my posts on the topic.  I like to wander.

                    2.  I find AOF’s posts fun and thought provoking.  He challenges norms of how I have seen finance presented elsewhere.  His writing makes me laugh out loud.  He combines finance with Avenger’s references…awesome!  Everything I read in finance doesn’t have to be actionable.

                    3.  He gets that math is not a crystal ball into the future.  Donnie’s, Zaphod’s, and Complete Newbie’s back and forth has reinforced to me that math is not a crystal ball.

                    4.  Worth vs. value.  Value is subjective, and I find it in his articles.

                     
                    Click to expand...


                    One of the big problems with math to describe anything is it gives us all a sense of security in the result, generalization or overall truth being presented...when in fact it may not be very good at all. Lulls us into belief or more commonly believing something stronger than we actually should.

                    Always easiest to fool ourselves. Doing a couple studies and publishing taught me everything I need to know about the reality of these things.

                    Comment


                    • #85
                      How to Lie with Statistics by Huff is a fun read if you haven't read it yet.

                      Comment


                      • #86





                         

                        However the insight was surprisingly banal. My conclusion was that a ~4% SWR was probably fine around 95% of the time.

                         

                         



                        This encapsulates my objection to what you did.  You were trying to improve the precision of the ~4% rule with a technique that can’t possibly provide such precision.  If you wanted to measure the importance of mean reversion to the 4% rule, there are better ways to do that.   Hopefully you get some eyeballs at least!

                         

                        Comment


                        • #87





                           

                           

                          Why I Did It?

                          My main interest here was to see how investment strategies might change if we drop any assumption on mean reversion. Since this is such visceral belief by individual investors I thought reproducing the Trinity Study through this lens might be provocative and insightful. I think it was both provocative and insightful – mission accomplished.


                          Click to expand...


                          That's exactly it. You looked at something that ignored mean reversion. I think the data, limited as it is, supports mean reversion fairly well, at least over reasonable time periods. So an analysis that ignores that might be theoretically interesting, but I would argue practically less than useful, especially given the conclusion, which basically supports what the analyses that did not ignore mean reversion said.

                          Evidence based investing is an interesting thing. I think it's a very good idea to look at the data, but you should always keep in the back of your mind just how little data there is and that none of it is the equivalent of a double blind, placebo controlled, prospective trial. So don't put too much faith in it.
                          Helping those who wear the white coat get a fair shake on Wall Street since 2011

                          Comment


                          • #88
                            Prospective analysis can only be done by simulation using MC models in finance. There is no double blind rondamized control trials so no use for that. This backtesting is still useful , infact you are sort of proving the point : if ignoring mean reversion yielded the same SWR, how useful is mean reversion as a factor? ( not much).

                            What you are saying is now is post hoc statements, easy to say now but it's proven with decent analysis by AoF (he already mentioned in his post here)

                            To each his own I suppose.

                            Comment

                            Working...
                            X