Apparently Johnny Depp is on the verge of bankruptcy. The company he is suing (that is counter suing) said he spent $2M a month for 20 years (a staggering almost half a billion dollars). He owns more 14 properties including a few islands on the Bahamas and a castle in France, 200 works of art, 45 deluxe cars (though he probably sold some along the way). I have no idea if his 'investments' are going to pay off (probably wont) but he is a great example of being asset rich but cashflow poor. Save some money Jonnhy!
Global Macro Trading Journal
- Thread starter Daal
- Start date
This Fed meeting was a non-event as it should have been. What matters is the next one and the ones after that, which have Fed forecasts with them. To me, what matters is the Fed's long-term forecast for:
-Fed funds (not the 'dots' which the media will all over but are actually irrelevant) but the longer run projection
-Long-run inflation and GDP
To me, this is the indicator of how the Fed is seeing the Trump policies. If they projected long run real interest rate rises, I might short stocks, bonds immediatly because that wouldn't be good at all. If the long-run projected GDP rises, that's a nod to Trump and stocks are a buy
-Fed funds (not the 'dots' which the media will all over but are actually irrelevant) but the longer run projection
-Long-run inflation and GDP
To me, this is the indicator of how the Fed is seeing the Trump policies. If they projected long run real interest rate rises, I might short stocks, bonds immediatly because that wouldn't be good at all. If the long-run projected GDP rises, that's a nod to Trump and stocks are a buy
Was playing around with the Kelly Criterion formula.
According to it, the optimal bet for 2-1 payoff trade with a 40% chance of winning is 10% of capital.
But what's interesting is that if you take the payoff and turn into a "subrime CDS" pay off level of 50 to 1, the optimal bet is 38.8%. And the bulk of this 'jump' happens from 2 to 1 to 6 to 1. At 6-1 the optimal bet is 30%. After that, you keep adding benefits to the trade (by increasing the pay off) but the bet size doesn't increase much. If you use half-Kelly (for further protection against estimation errors) then the bet size range is between 5% (2-1) to 15% (6-1)
Even at a 99-1 pay off, the bet size is still 39% (full kelly) or 18% (half kelly).
If you put a 99-1 pay off and an optimistic 75% win rate, the bet size comes out at 74.75% of capital, which at half kelly is 37.37%
Most of the best risk managers out there wouldn't recommend going above the half kelly, the world is too uncertain and people are too overconfident. To consistently do that is to leave one very vulnerable to negative long-term returns (despite postive expected value bets).
If you took all the people in the Forbes 400 and looked at their bet sizes from that perspective, how many of them would have stayed within the limits of what a good risk manager would suggest? Most of the really rich folks are people that bet all their money in 1 company and stayed with that for most of their life. And that company turned out to be a 100 bagger or more, or something like that. According to math and common sense, there is almost no situation where one should risk half their money, and yet, business people routinely risk everything. Does that mean they are idiots? I think it does suggest they are reckless but perhaps, are they also considering future income (from a job that they would have to get in case they failed) as part of their 'betting stake'? They know they probably won't starve to death if they failed, so some sort of safety net they are using in order to risk so much and still be comfortable with those oversized bets
According to it, the optimal bet for 2-1 payoff trade with a 40% chance of winning is 10% of capital.
But what's interesting is that if you take the payoff and turn into a "subrime CDS" pay off level of 50 to 1, the optimal bet is 38.8%. And the bulk of this 'jump' happens from 2 to 1 to 6 to 1. At 6-1 the optimal bet is 30%. After that, you keep adding benefits to the trade (by increasing the pay off) but the bet size doesn't increase much. If you use half-Kelly (for further protection against estimation errors) then the bet size range is between 5% (2-1) to 15% (6-1)
Even at a 99-1 pay off, the bet size is still 39% (full kelly) or 18% (half kelly).
If you put a 99-1 pay off and an optimistic 75% win rate, the bet size comes out at 74.75% of capital, which at half kelly is 37.37%
Most of the best risk managers out there wouldn't recommend going above the half kelly, the world is too uncertain and people are too overconfident. To consistently do that is to leave one very vulnerable to negative long-term returns (despite postive expected value bets).
If you took all the people in the Forbes 400 and looked at their bet sizes from that perspective, how many of them would have stayed within the limits of what a good risk manager would suggest? Most of the really rich folks are people that bet all their money in 1 company and stayed with that for most of their life. And that company turned out to be a 100 bagger or more, or something like that. According to math and common sense, there is almost no situation where one should risk half their money, and yet, business people routinely risk everything. Does that mean they are idiots? I think it does suggest they are reckless but perhaps, are they also considering future income (from a job that they would have to get in case they failed) as part of their 'betting stake'? They know they probably won't starve to death if they failed, so some sort of safety net they are using in order to risk so much and still be comfortable with those oversized bets
But of course, a lot of the time these people can't sell because it would spook private investors if the CEO is just dumping stock like there is no tomorrow. So its possible that there are forced to be on their hugely sized bet. When the stock goes public, then they can sell more easily but as a stock investor, its important to take this into account when one sees insider sales. Any rudimentary version of a position sizing model would tell a person with 80-90% of their networth in 1 stock to sell
But one thing that I don't get it why did Buffett own UST bonds in his private account in the years leading up to 2008 and then in 2008, he switched from bonds to stocks, apparently for up to 100% of his allocation
http://www.nytimes.com/2008/10/17/opinion/17buffett.html
"So ... I’ve been buying American stocks. This is my personal account I’m talking about, in which I previously owned nothing but United States government bonds. (This description leaves aside my Berkshire Hathaway holdings, which are all committed to philanthropy.) If prices keep looking attractive, my non-Berkshire net worth will soon be 100 percent in United States equities."
If you are really rich, it makes all the sense in the world to have some money stashed away just in case there is some kind of black swan/tail event and you lose it all (or most of it).
That's what I think he was doing by having some US bonds outside Berkshire.
What is not consistent with that is to have evidence that the black swan is showing up (2008 crisis), then take the money that you had stashed away (that would guarantee you would be wealthy no matter what happened) then put it in something highly correlated to what your main assets are in. Essentially, Buffett shorted even more puts on the thesis that 'America always comes back', and he put even his safe cash in there. He is basically saying that the utility of money doesn't matter to him, risk of ruin doesn't matter to him, all that he cares about are benefits
But as I have shown, adding benefits to a bet sizing formula, does not increase the recommended bet all that much (after a point). And if one is using robust assumptions (considering that the world is uncertain and we can be wrong all the time), putting 100% of his networth (or close to it) in the short put 'america always comes back' trade just looks completely reckess, even though it worked. Those buys could have been buys from the folks that bought US stocks in 1931. It gets worse when you consider that Berkishire is leveraged 1.2-1 or more
Is this man tail risk blind?
http://www.nytimes.com/2008/10/17/opinion/17buffett.html
"So ... I’ve been buying American stocks. This is my personal account I’m talking about, in which I previously owned nothing but United States government bonds. (This description leaves aside my Berkshire Hathaway holdings, which are all committed to philanthropy.) If prices keep looking attractive, my non-Berkshire net worth will soon be 100 percent in United States equities."
If you are really rich, it makes all the sense in the world to have some money stashed away just in case there is some kind of black swan/tail event and you lose it all (or most of it).
That's what I think he was doing by having some US bonds outside Berkshire.
What is not consistent with that is to have evidence that the black swan is showing up (2008 crisis), then take the money that you had stashed away (that would guarantee you would be wealthy no matter what happened) then put it in something highly correlated to what your main assets are in. Essentially, Buffett shorted even more puts on the thesis that 'America always comes back', and he put even his safe cash in there. He is basically saying that the utility of money doesn't matter to him, risk of ruin doesn't matter to him, all that he cares about are benefits
But as I have shown, adding benefits to a bet sizing formula, does not increase the recommended bet all that much (after a point). And if one is using robust assumptions (considering that the world is uncertain and we can be wrong all the time), putting 100% of his networth (or close to it) in the short put 'america always comes back' trade just looks completely reckess, even though it worked. Those buys could have been buys from the folks that bought US stocks in 1931. It gets worse when you consider that Berkishire is leveraged 1.2-1 or more
Is this man tail risk blind?
Last edited:
I was able to locate UK data to run some portfolio tests on. Unfortunately its annual data which creates the effect of making stocks look more stable and less volatile than they are but still, I can learn some things from it regardless
Some things that I was able to learn by looking at the data (preliminary findings that might change if I notice errors):
-The UK had a much less bad experience in the Great Depression, with annual data, the worst drawdown in the stock market was -23% real, vs -54% in US stocks. The decision from the UK to get out of the gold standard in 1931 vs the US in 1933 look like helped to prevent the deflationary spiral from getting worse. As a result stocks recovered a lot faster. Returns adjusted by the local (UK) price index and the GBP rate (in the case of gold):
Returns adjusted by the local (US) price index:
-UK bills beat the UK gilts 40% of the years
-UK bills beat UK stocks 27% of the years
-The UK yield curve was inverted (UK bill returns were higher than the yield on UK gilts) ~12% of the years
-UK Gilts had many staggering drawdowns. During WW1 they lost -66% (total real return) from 1914 to 1920 (The war was over after 1918 but there was probably some money printing going on to pay for things after it).
US did a little better, my UK gilts data is with 20y maturities, my US data is for 10y maturities. Total real return in US bonds in WW1 (plus 2 years after) was -34%, if I do a rough adjustment for the duration difference, I'm getting to -51% real. A little less bad
Gilts also lost -43% real from 1972 to the end of 1974 during their fiscal crisis when they had to be bailed out by the IMF. Vs -14% real in US bonds, or -21% duration adjusted
On the other hand, US bonds got hurt in the late 80's with a -34% real drawdown vs +9% real from UK gilts.
The UK might offer a guideline of what to expect with the US, if it continues on its path of deficits and war spending. We can expect a fiscal crisis with a big loss to holders of US bonds (especially for holders with a lot of duration), inflation and a collapse in the exchange rate
Some things that I was able to learn by looking at the data (preliminary findings that might change if I notice errors):
-The UK had a much less bad experience in the Great Depression, with annual data, the worst drawdown in the stock market was -23% real, vs -54% in US stocks. The decision from the UK to get out of the gold standard in 1931 vs the US in 1933 look like helped to prevent the deflationary spiral from getting worse. As a result stocks recovered a lot faster. Returns adjusted by the local (UK) price index and the GBP rate (in the case of gold):
Returns adjusted by the local (US) price index:
-UK bills beat the UK gilts 40% of the years
-UK bills beat UK stocks 27% of the years
-The UK yield curve was inverted (UK bill returns were higher than the yield on UK gilts) ~12% of the years
-UK Gilts had many staggering drawdowns. During WW1 they lost -66% (total real return) from 1914 to 1920 (The war was over after 1918 but there was probably some money printing going on to pay for things after it).
US did a little better, my UK gilts data is with 20y maturities, my US data is for 10y maturities. Total real return in US bonds in WW1 (plus 2 years after) was -34%, if I do a rough adjustment for the duration difference, I'm getting to -51% real. A little less bad
Gilts also lost -43% real from 1972 to the end of 1974 during their fiscal crisis when they had to be bailed out by the IMF. Vs -14% real in US bonds, or -21% duration adjusted
On the other hand, US bonds got hurt in the late 80's with a -34% real drawdown vs +9% real from UK gilts.
The UK might offer a guideline of what to expect with the US, if it continues on its path of deficits and war spending. We can expect a fiscal crisis with a big loss to holders of US bonds (especially for holders with a lot of duration), inflation and a collapse in the exchange rate
One thing that I'm finding is that the effects of the wars, fiscal problems, etc rendered UK gilts bad 'hedge assets'. I used Excel Solver to try the ideal allocation to my UK data, it was quite a bit different from the US case. The computer was forced to run to gold because UK gilts didn't produce as good as a downside cushion effect as US bonds did in a US centric portfolio.
So far, in my preliminary findings, this was the best risk/reward portfolio (as measured by both Sharpe and Sortino ratios) I could find:
Stocks tend to be overweighted in annual data because of the smoothing of drawdowns that tend to happen. For that reason, I don't think annual data is all that reliable and that's not what I'm trying to draw conclusions from. I'm trying to see if there are differences between UK ideal portfolios and US ideal portfolios. With UK data the best Sharpe and best Sortino congerved to the same thing, in the US the ideals were different:
I think this raises on important risk point. If the US continues down its road of fiscal deficits, balooning debt, lack of entitlement reform and wars, US bonds won't work as a 'hedge asset' as well as it did in the past. Gold (and/or global bonds) are likely to be superior hedges
Anoyone being guided by Buffett's 'perma' thesis that "America always comes back" needs to incorporate risk factors such as this one that will render historic experience quite different from the past
So far, in my preliminary findings, this was the best risk/reward portfolio (as measured by both Sharpe and Sortino ratios) I could find:
Stocks tend to be overweighted in annual data because of the smoothing of drawdowns that tend to happen. For that reason, I don't think annual data is all that reliable and that's not what I'm trying to draw conclusions from. I'm trying to see if there are differences between UK ideal portfolios and US ideal portfolios. With UK data the best Sharpe and best Sortino congerved to the same thing, in the US the ideals were different:
I think this raises on important risk point. If the US continues down its road of fiscal deficits, balooning debt, lack of entitlement reform and wars, US bonds won't work as a 'hedge asset' as well as it did in the past. Gold (and/or global bonds) are likely to be superior hedges
Anoyone being guided by Buffett's 'perma' thesis that "America always comes back" needs to incorporate risk factors such as this one that will render historic experience quite different from the past
This hypothesis seems to be confirmed further by this test
A 50% stock index position with 20% in gold, is better off (from a risk adjusted point of view) by incorporating UK bills into the rest of the portfolio than by adding UK gilts. UK gilts historically have been 'bad' hedge assets. Of course, by using bills, one sacrifices some returns, but one also avoids big ocasional drawdowns.
I think this is all very interesting because it will be a problem I will have to face when builing an ideal Brazilian portfolio. Duration can be a great hedge, it also can blow in your face quite badly and be a bad hedge. Given that most governments in the world don't borrow for long periods of time (how much of that is due preference vs inability is an interesting question that I do not have the answer to) one could argue that duration should be treated as a risky (or semi risky) asset. Or to coin a new term, a "risky hedge"
A 50% stock index position with 20% in gold, is better off (from a risk adjusted point of view) by incorporating UK bills into the rest of the portfolio than by adding UK gilts. UK gilts historically have been 'bad' hedge assets. Of course, by using bills, one sacrifices some returns, but one also avoids big ocasional drawdowns.
I think this is all very interesting because it will be a problem I will have to face when builing an ideal Brazilian portfolio. Duration can be a great hedge, it also can blow in your face quite badly and be a bad hedge. Given that most governments in the world don't borrow for long periods of time (how much of that is due preference vs inability is an interesting question that I do not have the answer to) one could argue that duration should be treated as a risky (or semi risky) asset. Or to coin a new term, a "risky hedge"