what should i be doing with surplus bankroll?

[Dazarath]

When the OP creates a thread asking "where to put excess bankroll", I think it's really inappropriate to give speculative answers such as uranium stocks/Iowa farmland(WTF)/GOOG/etc. A lot of the time, these people's don't have the time to research specific stocks, and a lot of the time they might not have the understanding to tell the difference between a helpful and a useless answer.

It should be painfully obvious that people asking these types of questions are looking for general answers, rather than "what is good to invest in right now?" This is why all of the other posters are suggesting things like high-yield savings accounts, S&P index funds, etc. Because these are decent answers all of the time.

For stuff like uranium stocks/Iowa farmland/GOOG, a lot of the time, it's just the poster speculating. Unless you actually believe that those are good investments all of the time, in which case you better have a good supporting reason. You're going to be hard pressed to convince anyone here of that.

If someone wants to speculate in uranium stocks/Iowa farmland/GOOG, they are completely free to do that. But I keep seeing people randomly tell others to buy these types of things, with no supporting reason other than "it's HOT HOT HOT!"

[Jeff W]

[ QUOTE ]
WITHOUT leverage, it is STILL possible to create a portfolio like the ones i drone on and on about. you just have to be creative and it has other risks inherent in it. for instance, instead of leveraging a 10yr 2:1, you can move out the yield curve to 30yr and increase duration. thus, your interest rate sensitivity will increase (though not by 2:1 but enough to give you a boost in returns). the risk here is yield curve risk. a 2:1 lev'd 10yr will do great if the mid section of the yield curve drops while the far end remains constant. thats a risk there.

[/ QUOTE ]

I don't think this would work... you get very little compensation for extending bond duration and you increase correlation with equities(reducing portfolio efficiency) when you do so.

I don't really care much about sharpe ratio... I don't think it's a great way to look at portfolios. Volatility is overrated as a risk metric.

[DcifrThs]

first off, i reread my posts and i come accross like a total ass. so for that i apologize. no excuse, just me being an ass for no real reason. i'm sorry.

in terms of my training, i worked at a top 3 hedge fund and was trained there. but i'm nowhere near where i want to be in terms of theoretical & practical knowledge.

to your post: i don't see why you can't just use money NOT used for long term retirement type investing in alpha and leave the rest for an optimal beta portfolio. lets say you have 100k. you have that allocated accross asset classes. you can reduce the amt of money needed to create that beta exposure via futures & other derivatives (total return swaps, repos etc.). that will leave you with some extra money to use for alpha. this is what is meant by the separation of alpha & beta. you don't need any extra money.

it sounds like you want to do the active management yourself so thats great. you also seem to have ideas that you are going with...if you want to improve your alpha portfolio construction ability, you can read up on portfolio structuring and some basic calcs to determine how much money you have at risk with every unit of movement of each position in your portfolio.

high returns though are only 1 part of it also. for both alpha and beta you (typically) want to maximize your returns while at the same time minimizing your risk. the reason that is important wrt alpha is simply that you want to be in a position to ADD to your exposure when the mkt moves against you (assuming you are still confident you have the bet right and nothing else has changed) rather than have to pare back exposures b/c you've taken too big of an exposure in the first place.

leveraging equities is fine, but equities already have embedded leverage (D/E>0). you'd be taking a risky security and making it even more risky. to do that (based on price moves alone) i think futures would be the way to go rather than options (which also have exposures to implied volatility and interest rates in addition to time vs. intrinsic value AND come with a premium...but, you don't have to post margin to hold an option whereas you do with a futures contract)

in terms of logistics of leverage: in order to increase the risk of bonds you can use futures, repos, and some combination of the physical bonds themselves to hit your target.

going back to the portfolio construction issue for a second. you said:

[ QUOTE ]
You sound like you have a better technical understanding of investing than I do...however I am spectical that your super-diversified REIT/bond/stock portfolio would turn a higher profit than a growth focused equity portfolio (given 30 years).

[/ QUOTE ]

thats fine if you dont care about risk. but the goal of investing shouldn't be maximum EV. it should be MaxEV/Min(Risk | Max EV) or another way Excess return (total return - cash or benchmark)/ standard deviation of excess return. that is called a sharpe ratio and this engineer type guy (i forget his field) Markowitz determined that, in theory, optimal portfolio construction should begin with the premise that you should try to maximize your risk adjusted returns (sharpe ratio). that is still true to this day. a higher sharpe ratio means that you have fewer years to be 95% confident your returns are past a certain threshold AND the length & breadth of your drawdowns will be significantly shorter & shallower.

the portfolio i'm talking about all the time accomplishes this goal. if you are unsatisfied with that level of risk/return, you can then leverage EVERYTHING proportionally to increase both yoru risk and return but still vastly outperform an equity only portfolio. BUT, if you can't use leverage (tax sheltered accounts) you do have to give up some level of return to maximize your sharpe ratio. thinking about this more, it seems that the difference between tax sheltered & taxable accounts is noticeable unless they let you invest in leveraged accounts. i guess one way around that would be to do the "target 2050" type thing and go very heavy on equities early on and then diversify as you get older (but not just into bonds). i'd need to think about this more and is probably a good thought exercise anyways.

i think i covered everything i n your post. if not, feel free to ask again.

Thanks,
Barron

[DcifrThs]

[ QUOTE ]
[ QUOTE ]
WITHOUT leverage, it is STILL possible to create a portfolio like the ones i drone on and on about. you just have to be creative and it has other risks inherent in it. for instance, instead of leveraging a 10yr 2:1, you can move out the yield curve to 30yr and increase duration. thus, your interest rate sensitivity will increase (though not by 2:1 but enough to give you a boost in returns). the risk here is yield curve risk. a 2:1 lev'd 10yr will do great if the mid section of the yield curve drops while the far end remains constant. thats a risk there.

[/ QUOTE ]

I don't think this would work... you get very little compensation for extending bond duration and you increase correlation with equities(reducing portfolio efficiency) when you do so.

I don't really care much about sharpe ratio... I don't think it's a great way to look at portfolios. Volatility is overrated as a risk metric.

[/ QUOTE ]

i think your first point is correct and i forgot about that. so wrt to tax sheltered accounts, seems like it is tough to get a more efficient portfolio with less risk at the same level of return.

volatility as a # isn't the best way to think about risk. what i mean is when you just take historical volatility or some #, you become far to dependent on that input. using a range though i do think it is a good way to think about portfolio risk. especially combined with logical thinking about what you'd expect volatility to be and not being a slave to what it was.

what would you use instead for portfolio construction? performance comparisons? etc.?

thanks,
Barron

[Jeff W]

[ QUOTE ]
volatility as a # isn't the best way to think about risk. what i mean is when you just take historical volatility or some #, you become far to dependent on that input. using a range though i do think it is a good way to think about portfolio risk. especially combined with logical thinking about what you'd expect volatility to be and not being a slave to what it was.


[/ QUOTE ]

There are a lot of risks not captured in volatility, even as a forward looking measure. For example, consumption risks... you shouldn't buy asset classes that are highly correlated with your income stream because you don't want them to tank at the same time. Example: poker players should not buy stock in Party Poker, Neteller, etc... but that applies on a much broader level as well and is the current argument from guys like Fama/French for why there is a persistent value premium.

Volatility doesn't capture kurtosis or skewness, either. This is one reasons why Hedge Funds can look very attractive--they provide "alpha", but you're often only being compensated for taking on huge kurtosis. "Picking up pennies in front of a steamroller." See Harry Kat's SSRN papers for more on this (and replicating hedge fund risk/reward profiles with mechanically traded futures baskets).

And I've just scratched the surface here...

[Mr. Now]

max,

Mr. Now reads through your posts here and agrees 100%. The poker has a ton of volatility; successful players are adequately bankrolled to acceptable risk of ruin, and banking the rest in highly liquid low-risk instruments. The reason is the high volatility (risk) in the high-reward poker component.

max is looking at the poker activity as part of a portfolio of instruments. max knows WTF he is talking about.

Mr. Now wonders if max might be willing to elaborate on his MPT view of poker bankroll-- in much greater detail-- on this thread. Mr. Now also wonders if Barron might provide commentary. Both of these men make posts consistent with deep knowledge of MPT & finance.

RoR Reference (poker component of portfolio):
http://www.traderscalm.com/rorhandiwork2.html

[DcifrThs]

[ QUOTE ]
[ QUOTE ]
volatility as a # isn't the best way to think about risk. what i mean is when you just take historical volatility or some #, you become far to dependent on that input. using a range though i do think it is a good way to think about portfolio risk. especially combined with logical thinking about what you'd expect volatility to be and not being a slave to what it was.


[/ QUOTE ]

There are a lot of risks not captured in volatility, even as a forward looking measure. For example, consumption risks... you shouldn't buy asset classes that are highly correlated with your income stream because you don't want them to tank at the same time. Example: poker players should not buy stock in Party Poker, Neteller, etc... but that applies on a much broader level as well and is the current argument from guys like Fama/French for why there is a persistent value premium.

Volatility doesn't capture kurtosis or skewness, either. This is one reasons why Hedge Funds can look very attractive--they provide "alpha", but you're often only being compensated for taking on huge kurtosis. "Picking up pennies in front of a steamroller." See Harry Kat's SSRN papers for more on this (and replicating hedge fund risk/reward profiles with mechanically traded futures baskets).

And I've just scratched the surface here...

[/ QUOTE ]

thinking about the future of a volatility # takes into account kurtosis for passive portfolios.

in these, the reason you diversify has to do with the fact that you've also diversified your kurtosis risk. not all the investments you expose yourself to will suffer from the same or similar tail risks (or risks of not approximating the normal distribution to be precise with my words, but in most cases we are talking about tails). this is one reason i haven't mention to diversify. good catch.

in terms of alpha portfolios, if tails are abnormally fat for an individual bet (albeit some times you don't know it), you should be holding far less of it and be aware of the triggers. thus you tweak the volatility & correlation #s to try to capture this. if those #s aren't sufficient to reduce your exposure to it, then you have to make changes to your system or override it after realizing this mistake. it isn't a science though. it's an art and extremely difficult. alpha generation is tough and the only way to try to mitigate this risk is to diversify your alpha allocations as well.

as an example, Texas Teachers & Missouri state pension funds have allocated a very small amt of money to equity tranches of CDOs & CLOs (the most risky parts of these securities. what this means is that if the debt instruments from which these packages are created default or go under, the investors in the bottom or equity tranches lose their money first). they have to be careful with these allocations and select based on managers, which they try to do. the effort they put into this could be misplaced since they could, instead, NOT invest in them and think about what their biggest risks are and what they may have missed with the rest of their alpha generating $.

specifically, these tranches are very risky and suffer huge kurtosis risk imo based on how they are created and the fact that we havne't seen what can happen to them (i briefly touched on why above). i agree with some commentators that pension funds have no business investing in these bottom tranches of debt.

in this case, they suffer from the problem you're talking about...using volatility as a # when making alpha allocation decisions, rather than thinking about everything as a whole).

[as a quick aside: it's tough for pension funds like that though since the returns they promise are enticing. the syndicating banks also do a good job selling them. if they perform well for a while though, the funds may increase their allocation and keep doing that until one day, some HUGE thing happens and BOOM they lose a ton]

i've neglected this part and made the mistake of getting lax with my wording (in terms of "logical thinking" in the statement you quoted-without delving into what is involved in that). you're right about volatility not being a good measure of risk and above is an example why.

so again, good catch.

but for this type of discussion, or us comparing managers or portfolios, the sharpe ratio is a pretty good proxy. you provided another reason to diversify your retirement type allocations: equities also suffer from kurtosis risk. what else should we use for this purpose if not the sharpe ratio?

Barron

[Fishhead24]

IF FOR THE NEXT TWO YEARS ONE WERE TO INVEST IN AN INDEX FUND ON THE NYSE instead of INVESTING IN IOWA FARMLAND, they will be making a huge, huge financial mistake.

-FH-

[DcifrThs]

[ QUOTE ]
max,

Mr. Now reads through your posts here and agrees 100%. The poker has a ton of volatility; successful players are adequately bankrolled to acceptable risk of ruin, and banking the rest in highly liquid low-risk instruments. The reason is the high volatility (risk) in the high-reward poker component.

max is looking at the poker activity as part of a portfolio of instruments.

[/ QUOTE ]

man, thats a good point. i don't have $ at risk at poker anymore but i'm soon to be helping a friend of mine and poker player structure his investments both 'cause he's a great kid and 'cause i'll enjoy it.

some off the cuff general thoughts on it:

mid road poker players have a really tough time banking on $.

1) they could suffer large swings (see the high limit posts about MASSIVE variance)

2) poker could be outlawed

3) the fish could leave (game could fall in popularity)

4) taxes could unexpectedly increase or some other unexpected thing could come about to make poker more costly.

but how do those things correlate to your active & passive investments? and how can a responsible poker player invest?

my first thought would be to diversify the non-bankroll related funds away from poker & NOT use them as a bankroll (imagine they're illiquid). this is different than the conventional thoughts about bankroll: that it includes ALL liquid cash.

sure, it includes all liquid cash if you don't care about retirement and are only maximizing EV (and your ability and means to play poker). you should put money away that you will never touch in addition to a hugely padded roll.

since poker has so many other risks though, risk averse players (i know that could be an oxy moron), or 'responsible' players shouldn't allocate as much money as they currently do to betting on stocks...or for alpha generation in general since that $ has a higher tail risk embedded anyways (ESPECIALLY if you only have a few bets going at a time).

'im tired of typing and thinking now though and need a break.

this is definitely the thread to discuss this thread though. anybody else have other thoughts??

thanks,
Barron

[DcifrThs]

[ QUOTE ]
max,

Mr. Now reads through your posts here and agrees 100%. The poker has a ton of volatility; successful players are adequately bankrolled to acceptable risk of ruin, and banking the rest in highly liquid low-risk instruments. The reason is the high volatility (risk) in the high-reward poker component.

max is looking at the poker activity as part of a portfolio of instruments. max knows WTF he is talking about.

Mr. Now wonders if max might be willing to elaborate on his MPT view of poker bankroll-- in much greater detail-- on this thread. Mr. Now also wonders if Barron might provide commentary. Both of these men make posts consistent with deep knowledge of MPT & finance.

RoR Reference (poker component of portfolio):
http://www.traderscalm.com/rorhandiwork2.html

[/ QUOTE ]

thanks for your kind words.

personally i wouldn't use MPT (modern portfolio theory...basically the use of sharpe ratio etc. to optimize portfolios. post-MPT is concerned more with loss risks rather than upside risks and is interesting also). i would use it for the funds the poker player chooses to set aside though with some additional thought about risks not related to the assets in general.

thanks again for your kind thoughts and i hope the above post helps.
Barron