20% annual returns over 40 years...interested?

[Gosso]

Gosso showed pretty well how a leveraged instrument is basically the expected leverage minus the borrowing rate (LIBOR+ ~0.90% in the case lf TLT) for the leverage bit.

Clive taught me.

Code: Select all

                3x          2x          1x
CAGR          14.65       11.96        9.58
Sharpe         0.69        0.79         1.2
Volatility     21.5        14.5         7.2
MaxDD         -60.9       -37.3       -12.5

Do your figures include 25% in cash?  My figures are a little different.  I've assumed a 3x33.3, rebalanced at 20/46 bands, and used daily data back to 1972.  Here's what I got (EDIT: This graph is wrong...I used the wrong data set for the S&P500...see two posts down for corrected graph):



Seems to me that this does well during low/negative real rates.  Here are the nominal total returns from 1972 to 2012:

       CAGR          STDEV       # of Rebalances
1x     11.2%         9.4%           12
3x     19.1%      25.6%         63

Nominal total returns from 1980 to 2000:

       CAGR          STDEV
1x     9.6%          9.0%
3x     8.9%          24.0%

[Gosso]

Oopsies...I f***ed up  For some reason I had copied over the S&P 500 data that had the 200-day SMA applied...that's why there was no spike down in 2008.  Here is the correct graph:

Total Nominal Return of $100 Investment - Rebalancing Bands of 20/46

Leverage based on daily movement, expense of 0.87% has been subtracted, borrowing cost of 3-Month T-Bill (times two for 3x) has been subtracted as well.
Source: FRED

Total nominal return from 1972-2012:

       CAGR      STDEV
1x     11.2%      9.4%
3x     16.9%      28.8%

Total Nominal Return from Feb 1977- July 2012:

       CAGR      STDEV
1x     10.9%      8.8%
3x     15.8%      27.3%

Total Nominal Return from 1980- 2000:

       CAGR      STDEV
1x     10.9%      8.8%
3x     9.6%      27.7%

I'm currently levered up to 2.5X on a vPP so I think I'm going to have to dial it back down once real rates go back up

I don't know...the levered play is looking a little bubbly.  I'm thinking the market is about to start tossing some serious volatility to shake off the levered players, my guess is after the elections.

[MediumTex]

Wonk,

Thanks for the update.

You don't visit us enough.

[Wonk]

Wonk,

Thanks for the update.

You don't visit us enough.

Aw, shucks MT.  Glad to know I was missed    I forgot how outrageously consuming a start up is...

Congrats to you and Craig on the new book!  Might have to find some time to read that.

Gosso & Slotline--thanks for the very cool info.

Related note, I think I get why leverage is so appealing to high-net worth individuals.  One would assume if you are a high net worth individual, you're probably high on an org chart somewhere.  For that to happen, you most likely appreciate the concept of leverage: in time, people, resources.  You can't do great things without it. 

I would think the problem is most people who employ massive leverage fail to assess risk properly, so many get wiped out pretty quickly.

[melveyr]

Wonk,

Branching off of what you said, another interesting thing is that many of the rich and the bankrupt often arrived at their outcomes by pursuing the same strategy: highly leveraged concentrated positions.

Only in hindsight does one look like a genius and the other a fool.

[jokerjoe]

Sorry for bringing up an old thread but it seems the most active one in which people discuss leverage.

If levering by borrowing on margin you are exposed to short-term interest rates, which is no longer counter balanced by having cash.  If you were to use S&P futures for the equity part you could free margin for the other components, bringing you to just under 1.33x leverage without borrowing.

To lever the other components, one could possibly use a synthetic long (long a call option, short a put).  You'd pay the spread and possibly some other cost I can't think of, but I think it'd be cheaper than borrowing and you'd be less exposed to interest rate moves.

[melveyr]

I am seriously considering just going cashless with my PP once I get settled into a steady job. Taking away the cash component is essentially the same as levering up the PP at the risk free rate, or more accurately not delevering the major return drivers (LTT, gold, and stocks).

I am still leery about the whole idea of buying a 30 year Treasury and then watering it down with T-Bills. I am still seeing 30 year treasuries as largely equivalent to a levered bet on for example the 10 year, but historically at a higher interest rate. You can see this as sharpe ratios go down on bond returns the farther out you go. You are getting implcit leverage at a cost higher than the risk free rate, so why would you delever it with the risk free asset?

[Pointedstick]

I find that using cash as the bulk of my former savings account that comprised an emergency fund works well because it allows me to integrate those funds explicitly into the portfolio concept. With your cashless PP, I presume you'd still want to build up/have a bunch of cash in such an account, no? You could go with your 3x33 PP and put that savings account money in T-bills instead just to keep them in separate mental buckets. Imagine that--a VP composed of T-bills! 

[melveyr]

I find that using cash as the bulk of my former savings account that comprised an emergency fund works well because it sllows me to integrate those funds explicitly into the portfolio concept. With your cashless PP, I presume you'd still want to build up/have a bunch of cash in such an account, no? You could go with your 3x33 PP and put that savings account money in T-bills instead just to keep them in separate mental buckets. Imagine that--a VP composed of T-bills! 

Yup I was thinking something very similar to that. The main difference from the vanilla PP would be that if my PP eventually got really big my emergency fund wouldn't automatically scale up with it. But yeah the liquidity factor for short term bonds / savings accounts is huge 

[stuper1]

Bringing up an old thread here.

Melveyr, what would your rebalancing bands be for the 3x33?

[melveyr]

Bringing up an old thread here.

Melveyr, what would your rebalancing bands be for the 3x33?

Well it depends how much you want momentum to play a role in your portfolio. The wider the bands, the more dominated your portfolio becomes by its winners. If they continue to do well, you will do better than if you had rebalanced sooner. Wider bands also have lower transaction costs and require less maintenance. This is such a personal thing I that I am not comfortable declaring any one band "best" but something like 20/46 seems reasonable to me.

It is also REALLY important to understand that the 3x33 portfolio would be hurt more than a traditional PP in a tight money recession. Cash is the only reliable asset to hold when the Fed over tightens, but luckily the benevolent central bankers usually don't punish the markets for very long.

I have flip flopped back and forth internally on whether not to go cashless with my PP. Right now I am doing the vanilla PP but with i-bonds as the cash. I see i-bonds as a free lunch so I am fine incorporating them into the PP. However, if you want a juiced up PP i still think removing cash would be the first step.

[MachineGhost]

I have flip flopped back and forth internally on whether not to go cashless with my PP. Right now I am doing the vanilla PP but with i-bonds as the cash. I see i-bonds as a free lunch so I am fine incorporating them into the PP. However, if you want a juiced up PP i still think removing cash would be the first step.

All that would do is lever up the PP's unequal risk exposure even more.  Now, I don't mean the similar short-term volatility that all three assets are displaying right now, but the long-term loss exposure.  It makes no sense to construct a portfolio based on short-term measurements when rebalancing or changes are only made in the long-term.

[rocketdog]

I have flip flopped back and forth internally on whether not to go cashless with my PP. Right now I am doing the vanilla PP but with i-bonds as the cash. I see i-bonds as a free lunch so I am fine incorporating them into the PP. However, if you want a juiced up PP i still think removing cash would be the first step.

You could also simply reduce the cash exposure to 10% and leave the rest at 3x30%.  That way you'd still have some dry powder on hand "just in case".

[melveyr]

I have flip flopped back and forth internally on whether not to go cashless with my PP. Right now I am doing the vanilla PP but with i-bonds as the cash. I see i-bonds as a free lunch so I am fine incorporating them into the PP. However, if you want a juiced up PP i still think removing cash would be the first step.

All that would do is lever up the PP's unequal risk exposure even more.  Now, I don't mean the similar short-term volatility that all three assets are displaying right now, but the long-term loss exposure.  It makes no sense to construct a portfolio based on short-term measurements when rebalancing or changes are only made in the long-term.

Are you saying to weight based off of long run vol or off of max DD or something?

Honestly though, I really think the PP is pretty balanced right now. If gold was twice as volatile as the other components I would most certainly underweight it. But they are all so close right now that extra tinkering in the strategic weights would be fruitless.

[MachineGhost]

Are you saying to weight based off of long run vol or off of max DD or something?

Honestly though, I really think the PP is pretty balanced right now. If gold was twice as volatile as the other components I would most certainly underweight it. But they are all so close right now that extra tinkering in the strategic weights would be fruitless.

I don't think long-run trailing volatility has any predictive power other than as a target for mean reversion, so I'd vote for MaxDD.  If rebalancing indeed captures volatility, it may be a mistake to weight it some other way or there may be a better way to weight it to better capture the volatility.  Inquiring minds want to know.

EDIT: I think we touched on this already.  The loss of the volatility capture would have to be made up by better risk-reward performance from whatever method is the replacement.  My gut feel is volatility is a red herring because it does not capture risk and the volatility capture premium is insignificant compared to the actual risks.  But as I am fond of pointing out, you have to factor in transaction costs and taxes when determining what is better on a net basis.  Monthly rebalancing was highly likely unprofitable to do in the past.

[Wonk]

For fun I decided to go back and check on this portfolio.  Smart Money rolled into Market Watch in the last year or two and the nice portfolio tracking software vanished along with it.  I went to Morningstar to do some tracking.  Here are the results:

$10,000

Funds:

SSO
UGL
UBT
SHY

01/2010-01/2014: $15878 (no rebalancing), +58.7%

01/2010-01/2014: $16794 (15/35 rebalance bands, annually), +67.9%

Annual returns, after rebalance:

2010: 21.1
2011: 28.0
2012: 8.9
2013: (0.7)
2014: 3.5 YTD

I would have thought a leveraged portfolio like this would have trouble in a down or sideways year.  It actually performed better than the 1x portfolio. 

[dragoncar]

Any idea what the volatility was?

[Wonk]

Any idea what the volatility was?

Unfortunately, no.  But I would imagine it would be roughly 2x the standard PP. 

[dragoncar]

Any idea what the volatility was?

Unfortunately, no.  But I would imagine it would be roughly 2x the standard PP.

So roughly the same sharpe ratio ... you could theoretically dial it up to S&P-level returns with lower volatility or S&P-level volatility with higher returns (assuming past performance is in any way indicative of future returns....)

[Wonk]

Any idea what the volatility was?

Unfortunately, no.  But I would imagine it would be roughly 2x the standard PP.

So roughly the same sharpe ratio ... you could theoretically dial it up to S&P-level returns with lower volatility or S&P-level volatility with higher returns (assuming past performance is in any way indicative of future returns....)

Yes, theoretically.  Likewise, if there were .5x funds of the same asset classes, one could assume half the return and half the volatility.  Just from watching the 2x funds operate for 4 years within this type of portfolio, you can see the value of non-correlated assets when operating as part of a whole.  The orthodox PP triggers rebalancing every 2-3 years, whereas the leveraged PP triggers a rebalance nearly once a year.  Capturing the value of that movement is important.  I think the leveraged funds could be a problem if there are ever large dislocations in the market causing major tracking error from the underlying indexes.  Otherwise, it looks like it operates as expected.

[dragoncar]

Unfortunately, no.  But I would imagine it would be roughly 2x the standard PP.

So roughly the same sharpe ratio ... you could theoretically dial it up to S&P-level returns with lower volatility or S&P-level volatility with higher returns (assuming past performance is in any way indicative of future returns....)

Yes, theoretically.  Likewise, if there were .5x funds of the same asset classes, one could assume half the return and half the volatility.  Just from watching the 2x funds operate for 4 years within this type of portfolio, you can see the value of non-correlated assets when operating as part of a whole.  The orthodox PP triggers rebalancing every 2-3 years, whereas the leveraged PP triggers a rebalance nearly once a year.  Capturing the value of that movement is important.  I think the leveraged funds could be a problem if there are ever large dislocations in the market causing major tracking error from the underlying indexes.  Otherwise, it looks like it operates as expected.

Cool... I'm pretty surprised that the daily synthetic leverage funds operate in this manner over large amounts of time.  However, I'm still convinced that it makes sense to first adjust the cash portion to adjust risk/reward since you get to directly hold the assets.  Maybe a small leveraged ETF PP in a tax-advantaged account (where you can rebalance more easily) would be a good VP.

[Gosso]

Annual returns, after rebalance:

2010: 21.1
2011: 28.0
2012: 8.9
2013: (0.7)
2014: 3.5 YTD

I would have thought a leveraged portfolio like this would have trouble in a down or sideways year.  It actually performed better than the 1x portfolio.

Wonk, something seems wonky with your 2013 numbers.  I track a 3x33% of the 1x, 2x, and 3x PPs using Google Finance.  Here are the 2013 results:

1x = -3.8%
2x = -8.3%
3x = -11.3%

All of the portfolios were rebalanced in mid-May, and include interest and dividends.  Maybe your portfolio rebalanced at a better time than mine?  I doubt SHY would make that much of a difference...

[Wonk]

Wonk, something seems wonky with your 2013 numbers.  I track a 3x33% of the 1x, 2x, and 3x PPs using Google Finance.  Here are the 2013 results:

1x = -3.8%
2x = -8.3%
3x = -11.3%

All of the portfolios were rebalanced in mid-May, and include interest and dividends.  Maybe your portfolio rebalanced at a better time than mine?  I doubt SHY would make that much of a difference...

I think it has to do with rebalancing. 

The 2x PP triggered a rebalance after 2010 & 2011, but not after 2012.  At the beginning of 2013, the stock portion was just under 30% of the portfolio so it was overweight.  With the strong year in stocks, it pulled the other assets up.  By the end of 2013, stocks represented 44% of the porfolio and triggered another rebalance.  Rebalancing was end of year only, so that might have an influence on the numbers as well.

[Gosso]

I think it has to do with rebalancing. 

The 2x PP triggered a rebalance after 2010 & 2011, but not after 2012.  At the beginning of 2013, the stock portion was just under 30% of the portfolio so it was overweight.  With the strong year in stocks, it pulled the other assets up.  By the end of 2013, stocks represented 44% of the porfolio and triggered another rebalance.  Rebalancing was end of year only, so that might have an influence on the numbers as well.

You're right.  That makes sense.

[modeljc]

I would have thought decay would have shown up in four years of returns?  The annual 15/35 rebalance seems to help.  Appears to have a 13.8% annual return.