tag:blogger.com,1999:blog-5621914599310831423.post4307594238573162400..comments2019-07-21T08:51:55.113-07:00Comments on The Retirement CafĂ©: Black Holes, the Higgs Boson and Retirement PlanningDirk Cottonhttp://www.blogger.com/profile/05616143752082768155noreply@blogger.comBlogger13125tag:blogger.com,1999:blog-5621914599310831423.post-3287402117320283092019-05-04T07:54:33.549-07:002019-05-04T07:54:33.549-07:00That is technically correct but I don't find i...That is technically correct but I don't find it useful. If you define a "solvable problem" as one in which there is one optimal solution that can be known in advance, then I would agree the retirement problem is unsolvable. We can't predict the future.<br /><br />However, we can use probabilities to help us make the best bets and give us a best chance of winning.<br /><br />Take backgammon or blackjack, for example. Both are games of luck <i>and</i> skill. The best players make the best bets for each move realizing that, while they can't be guaranteed a victory, they will win more games in the long run this way.<br /><br />If the dice (or the deal) are unfortunate, we will likely lose both games no matter how skilled we may be. On the other hand, we can lose with good rolls if we make poor bets, like splitting a pair of tens or leaving an open stone that can be hit by rolling a seven. Lastly, we can win sometimes, though rarely, by doing something unwise. I watched a friend who didn't know better once split a pair of tens and win both hands. Similarly, retirement is a game of skill and luck. <br /><br />Make good bets and you are more likely (though not guaranteed) to achieve better results.<br /><br />So, if you're looking for the optimal way to plan your individual retirement in advance, you're wasting your time. Your time will be better spent learning the best bets and understanding that you can't win them all. But if you combine your bets wisely, you can win enough of them.<br /><br />You could similarly state that backgammon is an "unsolvable" game in the sense that there is no guaranteed optimal strategy. But good players will beat your pants off over the course of many games. That's because they're skilled.<br /><br />Thanks for the comment.Dirk Cottonhttps://www.blogger.com/profile/05616143752082768155noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-75694497090205955252019-05-04T07:29:53.107-07:002019-05-04T07:29:53.107-07:00I don't know FireCalc. I looked at it briefly....I don't know FireCalc. I looked at it briefly. I seem to recall that showing unpredictability is claimed as a strong point.<br /><br />I don't know if FireCalc estimates probability of ruin but I would caution the use of any model that does. Probability of ruin is not a robust metric and can change significantly just by running the model again and changing nothing but the random number draw. In other words, you can run the model 100 times and get significantly different results each time.<br /><br />I suspect that models that generate only one probability of ruin do so because they were only run with one draw. That's not Monte Carlo.<br /><br />A retirement model is not a retirement plan; it's just one tool. I suspect most retirees would be better off with a good retirement advisor. A good one is hard to find, IMHO, but worth the effort.<br /><br />Thanks for writing!<br /><br />Dirk Cottonhttps://www.blogger.com/profile/05616143752082768155noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-25393080530389855482019-05-03T20:14:29.061-07:002019-05-03T20:14:29.061-07:00I like FireCalc also, but for the opposite reason ...I like FireCalc also, but for the opposite reason of user 96Trees: it opened my eyes to how UNPREDICTABLE the outcome is. Simply leave the 3 default numbers in there, click submit, and you have an outcome range of anywhere from running out of money in 25 years, to almost 6 times the original portfolio value in 30 years. I was shocked at how wide a range it was. Since then I've read everything I can get my hands on about retirement planning, and have learned how complex it is (and there are a lot of moving parts). My wife & I decided to use a human planner, and we like the plan he has come up with. justlookinghttps://www.blogger.com/profile/13929731068916306258noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-1080804642844656282019-05-01T19:04:08.632-07:002019-05-01T19:04:08.632-07:00
Retirement planning is an unsolvable problem with...<br />Retirement planning is an unsolvable problem with unlimited variables!!<br /><br />Michael Rossnoreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-2609350944767106332019-05-01T15:23:24.033-07:002019-05-01T15:23:24.033-07:00As I said, building confidence in your plan is a g...As I said, building confidence in your plan is a good reason to use models.<br /><br />As for "accuracy" and "errors", I can guarantee that the models will be wrong. All models are wrong; some are useful. Or, to quote Robert Merton, "All models are an abstraction of complex reality. As an abstraction, the model is incomplete."<br /><br />I was questioning your statement that the models give you an idea of where you will be. I don't think they do. I think they create many scenarios, some that might be something like where you will be, but you only get one retirement so at best you will end up in one of them. It could be the best, the worst, somewhere in between, or in one that wasn't simulated. It's good to know the range of possibilities.<br /><br />I think that using lots of models and studying the output of each is a good strategy. I only worry that many people think they provide "the answer."<br /><br />Best. . .Dirk Cottonhttps://www.blogger.com/profile/05616143752082768155noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-26931783096218643552019-05-01T09:51:37.726-07:002019-05-01T09:51:37.726-07:00By errors I'm thinking of inflation rates and ...By errors I'm thinking of inflation rates and average returns, as two examples of where we can guess wrong. When I did a retirement budget I didn't figure I'd retire to a house with expensive propane instead of electric and wood heat. I underestimated my Medicare costs. In my favor I made some good real estate investments later on while buying retirement property. These are the kinds of errors I'm talking about. In the end I must have used a dozen retirement calculators, plus a couple of humans associated with my accounts. All to give me some confidence that I was going to be okay. 96Treeshttps://www.blogger.com/profile/08716538387188852143noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-26921729658505618052019-05-01T07:55:10.608-07:002019-05-01T07:55:10.608-07:00Looking for confidence (or the lack thereof) is a ...Looking for confidence (or the lack thereof) is a valid use of models even if they can't predict where you will end up. That's generally why I use them. <br /><br />What do you mean by "accurate"?<br /><br />Dirk Cottonhttps://www.blogger.com/profile/05616143752082768155noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-92225638702262884832019-05-01T07:47:35.835-07:002019-05-01T07:47:35.835-07:00Apologies, that response got published before I in...Apologies, that response got published before I intended. I wanted to add that none of the computer models I have used replace a good human retirement planner.Dirk Cottonhttps://www.blogger.com/profile/05616143752082768155noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-18221770490683991242019-05-01T07:43:49.560-07:002019-05-01T07:43:49.560-07:00All models are wrong. I think MC models provide mo...All models are wrong. I think MC models provide more information than spreadsheet models but so long as you understand a model and what it does well and poorly, and what you are trying to learn, you'll be OK. You especially need to understand that no model is predictive but that might not be your goal.<br /><br />Ken Steiner uses <a href="https://howmuchcaniaffordtospendinretirement.blogspot.com/" rel="nofollow">a spreadsheet model</a> to estimate the "safe" amount to spend in the current year. For that purpose, I'm not sure any other model is better, though I particularly like Ken's for that purpose.<br /><br />Steve Vernon argues that RMD is a good method for determining current safe spending. I believe part of his argument is that if you can't determine which model is better, you might as well select the easiest to use.<br /><br />There is <a href="https://www.bogleheads.org/forum/viewtopic.php?t=97352&start=600" rel="nofollow">a more extensive spreadsheet </a> at Bogleheads that I would recommend.<br /><br />It is not my intent to disparage any particular type of model. I vastly prefer MC models for exploring possible future outcomes, though I understand that I can't predict which outcome, if any, will describe my future. I don't like MC models that try to minimize probability of ruin because that is a flawed metric. Unfortunately, most do. I also prefer MC because it allows me to model the interplay of several variables simultaneously.<br /><br />I particularly like <a href="maxifi.com" rel="nofollow"> MaxiFi.com</a> because it uses MC but not probability of ruin (it maximizes standard of living) but it isn't free. BTW, I believe it would fit your needs quite nicely, including solving the tax problem you describe. It has the benefit of being developed by a top economist, Laurence Kotlikoff, and has been around a long time, both of which should increase your confidence in the model.<br /><br />You mention that retirement planning must consider "enormous variables." True, and we fit most of those variables with WAGs. I sense that many believe that because the answer comes from Monte Carlo simulation and a computer that it must be "right." <br /><br />Rather than say that I disparage spreadsheet models, I prefer to say that I am highly skeptical of all models and would recommend that they not be your sole source for a retirement plan because, as I said, they're all wrong but some are useful.<br /><br />Thanks for writing!Dirk Cottonhttps://www.blogger.com/profile/05616143752082768155noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-29022720242046026872019-04-29T19:24:07.872-07:002019-04-29T19:24:07.872-07:00Dirk, or anyone else for that matter. I'm look...Dirk, or anyone else for that matter. I'm looking for a calculator. The spreadsheet model you disparage. You have to start somewhere before starting to consider the enormous variables that undermine our planning. A tool of this nature may have been discussed in a prior post. If so, let me know. <br /><br />I'm having trouble trying to describe what I'm looking for, but here it goes. <br /><br />I'm trying to calculate the money needed from savings to fund what Social Security doesn't provide. <br />Assumption 1- Social Security doesn't change (not likely)<br />Assumption 2- Inflation remains the same (not likely)<br />Assumption 3- You don't want to read the legion of assumptions that must be made for a spreadsheet model.<br /><br />I know the theoretical amount for year 1, but that amount has to increase annually to compensate for the annual inflation rate. In addition, that greater amount is taxed and one must increase the withdrawal of savings to compensate for the loss to state and federal income tax. It would also be helpful if it could include the number of years I expect to be alive.<br /><br />Then I'll need to try to guess my RMD on an amount that may not be right that I may or may not has as part of my retirement.<br /><br />While not the same a shotgun fitting, there are similarities. When you adjust one variable, another variable changes.Michael Rossnoreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-74294087412993725492019-04-29T18:40:41.498-07:002019-04-29T18:40:41.498-07:00With any calculator how do you know the result is ...With any calculator how do you know the result is accurate. I was just looking for confidence that I was on the right track.96Treeshttps://www.blogger.com/profile/08716538387188852143noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-23887338313677171822019-04-29T11:30:35.832-07:002019-04-29T11:30:35.832-07:00How does it provide a "kind of a general idea...How does it provide a "kind of a general idea of where you will be"?<br /><br />It shows a huge range of possible outcomes that you <i>might</i> experience if you retired 50 times. You will retire once, so where do you believe you will generally be within that range?Dirk Cottonhttps://www.blogger.com/profile/05616143752082768155noreply@blogger.comtag:blogger.com,1999:blog-5621914599310831423.post-87087671330356668112019-04-19T12:23:25.638-07:002019-04-19T12:23:25.638-07:00An interesting calculator is Firecalc. It runs ab...An interesting calculator is Firecalc. It runs about 50 simulations and then draws a line graph of your future earnings. Actually it draws 50 lines on the graph so what you get is kind of a general idea of where you will be.96Treeshttps://www.blogger.com/profile/08716538387188852143noreply@blogger.com