Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Motronic
Nov 6, 2009

A Proper Uppercut posted:

I honestly didn't even get the info, I just heard "managed" and "no matching", I just said no.

Also, I should say, I'm not ultra dedicated to contributing to my retirement savings, which I know is a sin here, but I would never get close to maxing out the 401k (I don't make enough to do that anyway).

It doesn't matter if you'll get close to making out the $19.5k max for the 401(k). What matters is if you'll exceed the $5500 max for an IRA. Because then your next best likely tax advantaged space is the 401(k), even if it's awful.

Are you actively saving for retirement at all? Do you have a plan?

Adbot
ADBOT LOVES YOU

A Proper Uppercut
Sep 30, 2008

Motronic posted:

It doesn't matter if you'll get close to making out the $19.5k max for the 401(k). What matters is if you'll exceed the $5500 max for an IRA. Because then your next best likely tax advantaged space is the 401(k), even if it's awful.

Are you actively saving for retirement at all? Do you have a plan?

My plan has always been stick what I can afford into a 401k. I was actively contributing until I left my previous job a few months ago. But yes, my yearly contributions could possibly exceed $5500 (but honestly not by much), so I may have to look into the 401k after all.

Quaint Quail Quilt
Jun 19, 2006


Ask me about that time I told people mixing bleach and vinegar is okay

A Proper Uppercut posted:

The place I'm at now has a crappy managed 401k with no matching, so I don't think I'm going use it. Is there any reason I shouldn't just roll this into an IRA so I can continue contributing? I'd like to just stick everything into a target date account so I don't need to worry about rebalancing.
I'm a bargaining employee (union) so they don't match where I work, they pay into our pension though.

Personally I'm going to make my own Vanguard simple IRA once I save a bit more. I have 10 weeks of pay saved currently for my emergency fund.

Screw employee plan commissions and fees unless they match obviously.

drainpipe
May 17, 2004

AAHHHHHHH!!!!
Do you foresee yourself in this job for a really long time? If you think you won't stay for more than 2-3 years, then taking advantage of the tax advantaged space is probably still preferable to lovely 401k options since you can roll it out when you leave.

Entropy238
Oct 21, 2010

Fallen Rib
Can someone explain to me why good asset allocation can lead to higher growth? I get all the diversity benefits in terms of mitigating risk, volatility etc ... I’m basically trying to understand why you wouldn’t go full stonks given they have the highest growth rates over long periods of time.

Is it something along the lines of:

- At different points in time different asset classes will outperform others. Some will outperform, others will underperform.
- For long periods of time, stonks might have low, no or minus growth.
- By having a more diverse asset allocation, you’ll be able to capture growth in different assets when stonks might not be on the up.
- By capturing that growth and maintaining your asset allocation proportions, you’ll get greater growth over time than had you invested in stonks alone?

Entropy238
Oct 21, 2010

Fallen Rib
Obviously large, mid and small caps, international vs non international are all their own asset classes for these purposes ....

pmchem
Jan 22, 2010


Entropy238 posted:

Can someone explain to me why good asset allocation can lead to higher growth? I get all the diversity benefits in terms of mitigating risk, volatility etc ... I’m basically trying to understand why you wouldn’t go full stonks given they have the highest growth rates over long periods of time.

Is it something along the lines of:

- At different points in time different asset classes will outperform others. Some will outperform, others will underperform.
- For long periods of time, stonks might have low, no or minus growth.
- By having a more diverse asset allocation, you’ll be able to capture growth in different assets when stonks might not be on the up.
- By capturing that growth and maintaining your asset allocation proportions, you’ll get greater growth over time than had you invested in stonks alone?

Have a look at this thread, I think it captures the general idea very nicely:

https://twitter.com/10kdiver/status/1264622958468726785?s=20

There are other much more detailed discussions on the subject, of course. But basically: it's about managing a risk profile, as your intuition suggests.

Entropy238
Oct 21, 2010

Fallen Rib
Ok thanks that’s super helpful.

Bremen
Jul 20, 2006

Our God..... is an awesome God

pmchem posted:

Have a look at this thread, I think it captures the general idea very nicely:

https://twitter.com/10kdiver/status/1264622958468726785?s=20

There are other much more detailed discussions on the subject, of course. But basically: it's about managing a risk profile, as your intuition suggests.

This thread made my head hurt.

GoGoGadgetChris
Mar 18, 2010

i powder a
granite monument
in a soundless flash

showering the grass
with molten drops of
its gold inlay

sending smoking
chips of stone
skipping into the fog
It's not necessarily a good metaphor, since it's a gambling strategy, and savvy investors do not have a "gambler's mindset", but it does illustrate the advantages of how exposure to all potential winners and losers will give a high likelihood of an Acceptable Outcome while shaving off the chances of the extraordinarily good and extraordinarily bad outcomes.

pmchem
Jan 22, 2010


GoGoGadgetChris posted:

It's not necessarily a good metaphor, since it's a gambling strategy, and savvy investors do not have a "gambler's mindset", but it does illustrate the advantages of how exposure to all potential winners and losers will give a high likelihood of an Acceptable Outcome while shaving off the chances of the extraordinarily good and extraordinarily bad outcomes.

Dismissing the Kelly Criterion as a gambling strategy is... really selling it short. It was developed for information theory and communication applications. But, it has wide applications to conservative investing and is applied by professional financial advisors, Wall Street, and people such as Warren Buffett and Charlie Munger. From the twitter thread,

https://twitter.com/10kdiver/status/1264623933669531648?s=20
https://twitter.com/10kdiver/status/1264623935401803776?s=20
https://twitter.com/10kdiver/status/1264623936811044864?s=20

from investopedia,
https://www.investopedia.com/articles/trading/04/091504.asp

from a CFA @ CFA Institute (with links to others),
https://blogs.cfainstitute.org/investor/2018/06/14/the-kelly-criterion-you-dont-know-the-half-of-it/

A discussion of it in the context of Buffett's advice for his wife,
http://sachinashanbhag.blogspot.com/2016/05/warren-buffett-90-10-portfolio-and.html

I mean, this is partly why target date funds rebalance regularly. Total bond funds aren't part of TDFs just to take in new money. A mantra of many investors is that "time in the market beats timing the market"; but, if being in the market is so great, why not just go pure stock if your horizon is 15+ years out? Why do TDFs have total bond funds in them at all before near-retirement age? The answer is to be found in ideas such as the Kelly criterion.

GoGoGadgetChris
Mar 18, 2010

i powder a
granite monument
in a soundless flash

showering the grass
with molten drops of
its gold inlay

sending smoking
chips of stone
skipping into the fog
Eh, I don't doubt it's been used to sell books. I think it's a reasonable enough metaphor for the benefits of diversification though!

pmchem
Jan 22, 2010


GoGoGadgetChris posted:

Eh, I don't doubt it's been used to sell books. I think it's a reasonable enough metaphor for the benefits of diversification though!

sell books, guide nearly $100b worth of stock market transactions in one fund for about three decades, you know, basically the same thing:
http://www.eecs.harvard.edu/cs286r/courses/fall12/papers/Thorpe_KellyCriterion2007.pdf (sec. 9)

pokeyman
Nov 26, 2006

That elephant ate my entire platoon.

pmchem posted:

I mean, this is partly why target date funds rebalance regularly. Total bond funds aren't part of TDFs just to take in new money. A mantra of many investors is that "time in the market beats timing the market"; but, if being in the market is so great, why not just go pure stock if your horizon is 15+ years out? Why do TDFs have total bond funds in them at all before near-retirement age? The answer is to be found in ideas such as the Kelly criterion.

I'm not seeing the connection between the Kelly criterion and target date funds. I assume you're claiming that it's used by target date fund managers to decide what proportion of the portfolio should go to bonds, so if not that's maybe where I got lost. Why would a Kelly portfolio would be preferred to one maximizing risk-adjusted returns for a target date fund?

pmchem
Jan 22, 2010


pokeyman posted:

I'm not seeing the connection between the Kelly criterion and target date funds. I assume you're claiming that it's used by target date fund managers to decide what proportion of the portfolio should go to bonds, so if not that's maybe where I got lost. Why would a Kelly portfolio would be preferred to one maximizing risk-adjusted returns for a target date fund?

That's a great question. Let's take this in parts: I am NOT claiming it is directly used, unaltered, by target date fund managers. Vanguard describes some of their TDF methodology here: https://www.vanguard.com/pdf/s167.pdf. They refer to the glide path as "a broadly diversified portfolio that engages in limited market-timing (Brinson, Hood, and Beebower, 1986; Davis, Kinniry, and Sheay, 2007)" (yes, by using a target date fund you are market timing). I do not immediately have access to the articles by Brinson or Davis to peek into the details. The Kelly criterion is a rather simple method for sizing bets from around 60 years ago, so, I would hope that in TYOOL 2020, target date fund managers have come up with something a bit more robust for their task.

You mention risk-adjusted returns. To get people on the same page, this generally refers to Sharpe ratio,
https://www.investopedia.com/terms/s/sharperatio.asp
or Sortino ratio
https://www.investopedia.com/terms/s/sortinoratio.asp

In either case, it's an objective metric. A measurement. So you need a strategy or method to maximize that metric. Kelly is one potential path to doing that (depending on the "bets" being evaluated). From the twitter thread's coin-flipping example, outcomes are:

https://twitter.com/10kdiver/status/1264622986901909505?s=20

vs.

https://twitter.com/10kdiver/status/1264622997823827968?s=20

Using the Kelly bets, you have lowered your expected returns, but you have GREATLY decreased the standard deviation of those returns. Compared to the all-in strategy, the Kelly results have a greater Sharpe ratio and Sortino ratio -- greater risk-adjusted returns, which is exactly what you asked about!

In fact, Kelly can be directly applied to stock portfolios using Sharpe or Sortino as objective metrics, as is done here (and in many other places, such as by Thorpe...): https://arxiv.org/pdf/2007.06460.pdf

The math behind portfolio optimization can get complicated quickly depending on the objective function you're trying to maximize (and on whether leverage is allowed). That's another reason to doubt that TDF managers used unaltered Kelly directly. See one example of that on bogleheads, here: https://www.bogleheads.org/forum/viewtopic.php?t=305919

What I DO claim is that the ideas TDFs actually use in practice are related to the idea behind the Kelly criterion, as they serve the same purpose: sizing investments to achieve better* returns over a sequence of events (* where "better" is defined by some function, possibly but not necessarily a risk-adjusted return metric). I would wager that if you traced back citations in the Brinson or Davis papers, at some point Kelly is cited.

Regardless, it's a very neat idea to try and understand! And it's just a starting point. Ideas such as Kelly which whose results are very easily verified for toy examples are a launching pad to understand "why good asset allocation can lead to higher growth?", which was the posted question that started this whole discussion. Then if you really want to go down the rabbit hole and have a strong math background, you start reading papers elsewhere or posts on bogleheads like the one I linked.

It's a very profound investment idea, really: I think this thread generally endorses target-date funds, but how and why do they provide better risk-adjusted returns? How can they be understood? Why not just go 100% stocks if you're not retiring until 2050?

Duckman2008
Jan 6, 2010

TFW you see Flyers goaltending.
Grimey Drawer
I appreciate everyone’s effort posts on these analysis strategies , but I get index funds so I don’t have to put that much thought into it. Lol.

Entropy238
Oct 21, 2010

Fallen Rib
Ok so here’s a question that’s slightly related to my previous one.

To get the growth benefits of diversification a la the Kelly Criterion, do you consider all of your investments as a whole or should you consider individual investments and their related risks discretely?

For example, after fulfilling all expenses, needs and reasonable wants, a couple will have a certain amount of money to save / invest each month. This could go on deposit or into pension payments, mortgage overpayments, after-tax equity investing etc ...

Should the money that goes into an index fund via your pension be viewed as a discrete investment that might grow more by apportioning small amounts in cash funds / bonds? Or can you look at the fact you’re making mortgage overpayments and keeping money on deposit as fulfilling a similar function to the money you otherwise might have put in bonds / cash funds as part of the overall amount of money you’re investing?

doingitwrong
Jul 27, 2013

Duckman2008 posted:

I appreciate everyone’s effort posts on these analysis strategies , but I get index funds so I don’t have to put that much thought into it. Lol.

Unless you have turned your money over to a target date fund, you still have to decide which indices and in what proportion you are investing.

For people trying to get a general sense of what the Kelly Criterion is about, the first 13 mins of this lecture describes the problem of looking at average expected returns across space vs across time. Because we live in time.

https://youtu.be/f1vXAHGIpfc

doingitwrong fucked around with this message at 13:24 on Dec 26, 2020

Orange DeviI
Nov 9, 2011

by Hand Knit
Sounds like overthinking it to me

Duckman2008
Jan 6, 2010

TFW you see Flyers goaltending.
Grimey Drawer

doingitwrong posted:

Unless you have turned your money over to a target date fund, you still have to decide which indices and in what proportion you are investing.

For people trying to get a general sense of what the Kelly Criterion is about, the first 13 mins of this lecture describes the problem of looking at average expected returns across space vs across time. Because we live in time.

https://youtu.be/f1vXAHGIpfc

please knock Mom! posted:

Sounds like overthinking it to me

:same:


Like, I’m not knocking anyone delving into it, but I just keep a ratio of Vanguard total stock, large, small , international and bond stocks. The larger majority of mine is in my work 401k, which I have 1 index option of large, small, international and bond, so I’m lucky, it’s pretty easy for me.

Quick gripe: all my work target date funds have a 0.6% ratio, and they have 2 of either type (large, small, etc). The index funds are 0.05%, the non index funds are 0.6%, and they of course don’t make it easy to tell the difference.

Motronic
Nov 6, 2009

Duckman2008 posted:

:same:


Like, I’m not knocking anyone delving into it, but I just keep a ratio of Vanguard total stock, large, small , international and bond stocks. The larger majority of mine is in my work 401k, which I have 1 index option of large, small, international and bond, so I’m lucky, it’s pretty easy for me.

But but....investing must be more complicated than that!

pmchem
Jan 22, 2010


Duckman2008 posted:

Like, I’m not knocking anyone delving into it, but I just keep a ratio of Vanguard total stock, large, small , international and bond stocks. The larger majority of mine is in my work 401k, which I have 1 index option of large, small, international and bond, so I’m lucky, it’s pretty easy for me.

So, the first sentence states that you have 5 funds to choose from. Just curious, how do you decide upon your asset allocation and rebalancing among those 5 funds?

quote:

Quick gripe: all my work target date funds have a 0.6% ratio, and they have 2 of either type (large, small, etc). The index funds are 0.05%, the non index funds are 0.6%, and they of course don’t make it easy to tell the difference.

yeah, I know Schwab does that. Annoying.

Duckman2008
Jan 6, 2010

TFW you see Flyers goaltending.
Grimey Drawer

pmchem posted:

So, the first sentence states that you have 5 funds to choose from. Just curious, how do you decide upon your asset allocation and rebalancing among those 5 funds?


I picked the generic Vanguard small cap, mid cap, total index, international , and I have some in a vanguard target date fund. My version of “fun” in savings is occasionally buying an ETF of the Growth fund (maybe once every 1-2 years). I rebalance once a year, but I mostly just try to let it hang out and do it’s thing.


For the record , I guess there’s so much I can knock someone for getting very in depth on the long term investing thread. Like, this is the threads to post the analysis you are posting.

Boot and Rally
Apr 21, 2006

8===D
Nap Ghost

pmchem posted:



https://twitter.com/10kdiver/status/1264622997823827968?s=20

Using the Kelly bets, you have lowered your expected returns, but you have GREATLY decreased the standard deviation of those returns. Compared to the all-in strategy, the Kelly results have a greater Sharpe ratio and Sortino ratio -- greater risk-adjusted returns, which is exactly what you asked about!


That table says the expected returns would be higher, not lower. The Kelly column cuts out the low returns and piles that probability into the high returns. Is 'lower' measured relative to something other than 'all in'?

E: It appears there are some extreme outliers in the billions category so in the 'all in' you're less likely to end up there, but I guess it is infinite money.

Boot and Rally fucked around with this message at 17:09 on Dec 26, 2020

zaurg
Mar 1, 2004

Duckman2008 posted:

Quick gripe: all my work target date funds have a 0.6% ratio, and they have 2 of either type (large, small, etc). The index funds are 0.05%, the non index funds are 0.6%, and they of course don’t make it easy to tell the difference.

Same here, probably a common gripe.

Related question, what’s the math or is there an online calculator to determine the long-term difference between expense ratios... let’s say in a 401k over the course of ~25 years? I’ve been going with the funds with smallest ratios but curious to see how much of a difference 0.05% is to 0.6% in actual account balance by the time withdrawals are made.

Grumpwagon
May 6, 2007
I am a giant assfuck who needs to harden the fuck up.

zaurg posted:

Same here, probably a common gripe.

Related question, what’s the math or is there an online calculator to determine the long-term difference between expense ratios... let’s say in a 401k over the course of ~25 years? I’ve been going with the funds with smallest ratios but curious to see how much of a difference 0.05% is to 0.6% in actual account balance by the time withdrawals are made.

https://www.begintoinvest.com/expense-ratio-calculator/

Chu020
Dec 19, 2005
Only Text
Math is pretty easy, just choose your expected annualized return and subtract from that the ER when you calculate how much the account will be worth in 25 years. Any standard investment calculator on Google would be fine for this. For example, with an expected annualized return of 7%, you'd compare 6.4% (ER of 0.6) vs 6.95% (ER 0.05). Incidentally, if you use those numbers with maxing out your 401k every year for 25 years assuming a $19.5k contribution per year, then you end up with $1204427 vs $1313560, or a difference of just under $110k. Now that doesn't take into account inflation and changes in contributions, but you could just use an inflation-adjusted rate of return (like 4%) as your starting point if you wanted to do it that way instead.

Space Gopher
Jul 31, 2006

BLITHERING IDIOT AND HARDCORE DURIAN APOLOGIST. LET ME TELL YOU WHY THIS SHIT DON'T STINK EVEN THOUGH WE ALL KNOW IT DOES BECAUSE I'M SUPER CULTURED.

Boot and Rally posted:

That table says the expected returns would be higher, not lower. The Kelly column cuts out the low returns and piles that probability into the high returns. Is 'lower' measured relative to something other than 'all in'?

E: It appears there are some extreme outliers in the billions category so in the 'all in' you're less likely to end up there, but I guess it is infinite money.

Yeah, looking at pure expected value can take you to some strange places.

The best example of this is the St. Petersburg paradox: imagine a hypothetical lottery where the payout starts at $2 and doubles every time a fair coin comes up heads, but you walk away with the pot once tails comes up and the streak is broken. So, if the coin comes up tails on the first flip, you get nothing. If you get a run of two heads then tails, it's $4. If you manage a run of 10 heads in a row before you bust out, you get $1,024.

Mathematically, the expected value of that lottery is infinite. The probability of a long streak is tiny in the same sense that other galaxies are far away, but the massive payouts make up for it.

In practice, if you want to be a millionaire, you'd be better off with a Powerball ticket, because the long tail of one-in-a-quadrillion odds to become a quadrillionaire (and so forth, up to any arbitrary power of two you pick) aren't actually that great.

pmchem
Jan 22, 2010


Space Gopher posted:

Yeah, looking at pure expected value can take you to some strange places.

The best example of this is the St. Petersburg paradox: imagine a hypothetical lottery where the payout starts at $2 and doubles every time a fair coin comes up heads, but you walk away with the pot once tails comes up and the streak is broken. So, if the coin comes up tails on the first flip, you get nothing. If you get a run of two heads then tails, it's $4. If you manage a run of 10 heads in a row before you bust out, you get $1,024.

Mathematically, the expected value of that lottery is infinite. The probability of a long streak is tiny in the same sense that other galaxies are far away, but the massive payouts make up for it.

In practice, if you want to be a millionaire, you'd be better off with a Powerball ticket, because the long tail of one-in-a-quadrillion odds to become a quadrillionaire (and so forth, up to any arbitrary power of two you pick) aren't actually that great.

yeah, that video linked above by doingitwrong is pretty good. it covers st. petersburg (20m in), ensemble averages (expected value) vs. individual trajectory time averages, the kelly criterion explicitly (18 minutes in), etc.

Ole Peters has done some really interesting work in ergodicity economics and strategies related to targeting geometric return of investments:
https://ergodicityeconomics.com/

Damn Bananas
Jul 1, 2007

You humans bore me
I feel stupid asking this but... At Vanguard I have two accounts: an IRA, where it actually calls itself an IRA in the account title, and another account that is joint owned with my spouse and invested in VTSAX and VFIFX. It just calls my second account “Mary and John Smith,” our names, but has no other descriptive words outright calling itself anything. I was intending to open a brokerage account when I made it at the advice of what to do with excess money, but I have no idea if what I created is actually a brokerage account or something else. Is that just what a brokerage account looks like on the Vanguard account page?

withak
Jan 15, 2003


Fun Shoe
It is.

edit: Mine looks like this:

withak fucked around with this message at 22:00 on Dec 26, 2020

pmchem
Jan 22, 2010


Entropy238 posted:

Ok so here’s a question that’s slightly related to my previous one.

To get the growth benefits of diversification a la the Kelly Criterion, do you consider all of your investments as a whole or should you consider individual investments and their related risks discretely?

For example, after fulfilling all expenses, needs and reasonable wants, a couple will have a certain amount of money to save / invest each month. This could go on deposit or into pension payments, mortgage overpayments, after-tax equity investing etc ...

Should the money that goes into an index fund via your pension be viewed as a discrete investment that might grow more by apportioning small amounts in cash funds / bonds? Or can you look at the fact you’re making mortgage overpayments and keeping money on deposit as fulfilling a similar function to the money you otherwise might have put in bonds / cash funds as part of the overall amount of money you’re investing?

I don't think this has been answered yet.

In theory, it's very attractive to manage your assets as a whole portfolio. There are interesting arguments to be made in favor of, say, tilting heavy in stocks and away from bonds if you have a mortgage:
https://earlyretirementnow.com/2016/11/02/why-would-anyone-have-a-mortgage-and-a-bond-portfolio/

It is conventional wisdom to take advantage of tax-efficient fund placement when possible (IRAs vs taxable accounts),
https://www.bogleheads.org/wiki/Tax-efficient_fund_placement

However, in reality, it's not so simple. Bonds come in tax-exempt flavors, some popular bond yields are lower than stock index dividends these days, you might sell the house, rebalancing is much more advantageous within an account than across accounts, different accounts may have different investment options available (e.g. your 401k vs. IRA), etc. There's no clean answer to your question, which is probably why nobody answered it yet.

Personally, I track my investments as a whole and make sure the whole picture reflects my desired allocation, but within each investable account I try to make it so that I can benefit from rebalancing if one asset outperforms another (cash is also an asset in this scenario). Not sure if I'd recommend that to someone else, though. Highly individual situation.

moana
Jun 18, 2005

one of the more intellectual satire communities on the web

pmchem posted:

Ole Peters has done some really interesting work in ergodicity economics and strategies related to targeting geometric return of investments:
https://ergodicityeconomics.com/
Isn't this the physicist who thinks he's a genius for pointing out really narrowly defined problems that everyone knows about already? He's hyped up by the same group of online people who revere Nassim Taleb and the rest of the popsci echelon, and as far as I can tell no economists actually take him seriously. I mean, the Kelly criterion stuff has been used for decades and decades, it's nothing new to point out that average expected return isn't the only factor to consider when investing.

Has he actually been mentioned in any journals that aren't his own website? I'd like a link if there's stuff he's proposing that's actually new and useful but I'm really, really skeptical of these kinds of "we're overthrowing all of economics!" bold claims.

pmchem
Jan 22, 2010


moana posted:

Isn't this the physicist who thinks he's a genius for pointing out really narrowly defined problems that everyone knows about already? He's hyped up by the same group of online people who revere Nassim Taleb and the rest of the popsci echelon, and as far as I can tell no economists actually take him seriously. I mean, the Kelly criterion stuff has been used for decades and decades, it's nothing new to point out that average expected return isn't the only factor to consider when investing.

Has he actually been mentioned in any journals that aren't his own website? I'd like a link if there's stuff he's proposing that's actually new and useful but I'm really, really skeptical of these kinds of "we're overthrowing all of economics!" bold claims.

Shrug, yes his background is physics, but you asked for a link to his work so here it is, it's easily googleable:
http://tuvalu.santafe.edu/~ole/publications.html

Physical Review Letters is an extremely high-profile journal:
https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.110.100603

Nature Physics isn't bad either:
https://www.nature.com/articles/s41567-019-0732-0

Doesn't hurt to co-author articles with Nobel laureates:
https://aip.scitation.org/doi/10.1063/1.4940236

I mean you can demean him if you want but I'd rather discuss the arguments and not the fact that he was trained in physics?? Wall Street firms snap up physics PhD's left and right so it's not too much of a leap for a physics person to do research in certain areas related to investment.

edit: I love how this thread has come full circle from dismissing Kelly as a gambling strategy to saying it's been used for decades in economics and is basically obvious

pmchem fucked around with this message at 00:11 on Dec 27, 2020

acidx
Sep 24, 2019

right clicking is stealing
With our companies dental plan, I have to pay the costs out of pocket, then file a claim and get reimbursed by the company. Can I keep the receipt from when I pay the cost out of pocket and use that to reimburse myself out of my HSA down the road even if I get reimbursed by the company as well?

H110Hawk
Dec 28, 2006

acidx posted:

With our companies dental plan, I have to pay the costs out of pocket, then file a claim and get reimbursed by the company. Can I keep the receipt from when I pay the cost out of pocket and use that to reimburse myself out of my HSA down the road even if I get reimbursed by the company as well?

Not legally for full face value but you're really unlikely to be caught. For the unreimbursed out of pocket costs yes.

moana
Jun 18, 2005

one of the more intellectual satire communities on the web

pmchem posted:

edit: I love how this thread has come full circle from dismissing Kelly as a gambling strategy to saying it's been used for decades in economics and is basically obvious
The thread is made up of people with different opinions, I am not the one who dismissed it as a gambling strategy. If you hate this thread so much then don't read it.

I don't doubt his work as a physicist, I doubt his work in economics as being this groundbreaking thing that his twitter followers claim it to be. All of these papers are published in physics journals, right? Why is he being ignored by every journal in economics and finance? I'm not doing a deep dive into his arguments because it just doesn't seem like anyone is taking him seriously, and I don't have the economic background to know why instantly. I don't think that's an unreasonable position for a lay investor to have. I don't want to spend a lot of time digging into something if I'm going to be digging into a pile of worthless poo poo, you know?

I like Early Retirement Now because, like in the article you linked, a lot of his ideas are 1. Supported by not just theory, but also backtesting in real world conditions and 2. Able to be implemented. So I guess the question for me is: what specific changes have you made in your long term investing strategy due to Peters' work and why?

Entropy238
Oct 21, 2010

Fallen Rib

pmchem posted:

I don't think this has been answered yet.

Intuitively, my brain is telling me that to benefit from the Kelly Criterion investments need to be viewed discretely to the extent their risks are also discrete. While risks to your portfolio and their overall impact need to be considered in the round, at a practical level if the only way the criterion can be implemented is by allocation strategies within the same account then it would seem to make sense to examine investments individually to ensure growth is being maximised appropriately.

To use a gambling analogy: a gambler might seek to make money from bets in blackjack, poker and on the roulette wheel. The overall goal of the gambler (his strategy) is to maximise returns on his bets while minimising unnecessary risk. While at a macro level the gambler might decide to allocate his money to different kinds of bets based on his risk tolerance / desired returns, the risks associated with each kind of bet exist in isolation from the others. In other words, if blackjack is equity investing and the gambler suddenly discovers card counting (the Kelly criterion), this discovery has no impact on the risks associated with making bets at the poker table (early mortgage repayments) or roulette wheel (deposit accounts). Although, it might affect his overall allocation to the different bets.

In that sense, the benefit of the criterion isn't really in the guidance it provides in determining asset allocation as such, but rather its effectiveness as a tool for making particular investments more effective by maximising returns through the removal of unnecessary risks.

My intuition might be completely wrong though!

Entropy238 fucked around with this message at 01:30 on Dec 27, 2020

Motronic
Nov 6, 2009

moana posted:

what specific changes have you made in your long term investing strategy due to Peters' work and why?

Answering this would take away from the barely discernable pattern opposition posting in here.

Adbot
ADBOT LOVES YOU

pmchem
Jan 22, 2010


moana posted:

The thread is made up of people with different opinions, I am not the one who dismissed it as a gambling strategy.

ya, I'm genuinely happy that others are willing to chime in with that view. The whole discussion around Kelly started with a poster asking "why good asset allocation can lead to higher growth?" and I think the thread has shown a nice example of that on a toy problem of position sizing. I wish other people with better answers would step up and help out with their answers, too, because there are certainly more modern answers than that 60+ year-old basic strategy -- but I'm not aware of many examples that are quite so simply demonstrated and easy to grasp (other than just pointing at backtesting).

Anyway, re: Peters specifically, I wasn't the person who originally posted his video and I never endorsed a particular investment strategy based on his views. I just enjoyed the video and think some of his work looks interesting. I don't know anything in particular about his interactions with the academic economics community. You asked for links to his work and questioned its rigor so I provided links to some publications in the area. If you want to check out someone putting ideas aligned with Peters' into action, there's a guy who has been blogging in some detail about it: https://breakingthemarket.com/stochastic-efficiency-is-real-and-its-spectacular/ . I am not that guy and do not endorse that strategy so don't take the link in that manner.

I'm glad you like the Early Retirement Now link. I haven't read much from that blog -- I ran across the mortgage bit while doing mortgage reading -- but I guess I should check out more of it.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply