Global Macro Trading Journal

[luisHK]

Gold miners on fire as well. I'm not listening to Yellen's speech, just checking newswires, anyone care to explain why financials and USD are down and gold/bonds/stocks are up upon the news of a sound economy and expected sustained rate hikes ?
I like how the market reacts, but can't say I understand...

 

[Daal]

Gold miners on fire as well. I'm not listening to Yellen's speech, just checking newswires, anyone care to explain why financials and USD are down and gold/bonds/stocks are up upon the news of a sound economy and expected sustained rate hikes ?
I like how the market reacts, but can't say I understand...

The 'hawkish' fed trade was crowded, lots of people were in it, when the Fed didn't surprise to the hawkish side (more or less coming in line or perhaps a bit dovish), lots of people headed to the exits.

 

[Daal]

IB will now pay 0.41% on USD balances over $10,000. Pretty soon, cash rates will start to get decent and it might even make sense to park some there and not have to buy bonds instead

 

[luisHK]

The 'hawkish' fed trade was crowded, lots of people were in it, when the Fed didn't surprise to the hawkish side (more or less coming in line or perhaps a bit dovish), lots of people headed to the exits.

Thanks, this makes sense considering the available information.

 

[Daal]

El-Erian on the Fed. I agree with all of it. This Fed will lead the market rather than follow, March was the start of that process. By mid-year markets will find this out and they wont be pleased. Right now the Fed is being soft because it doesn't want a premature fight with Trump or to tank markets too much. But as the upside risks continue to play out, they will be forced. A 'soft' Fed will dissapear once inflation risks start to appear. But people don't get it, because they are not used to it. Its been a long-time but they will have no choice. Removing 'only' from the statement yesterday is part of that process, they don't want to be blamed for a shock if down the line they hike .50

 

[Daal]

I was thinking about the Barbell strategy and my portfolio tests and I decided to do some other things. I run a computer simulation for almost every asset mix possible but limiting the computer choices to only 'stock+bills' 'stock+gold' 'stock+bonds' then 'stock+gold+bills(no bonds)' then everything. and told the computer to tell me the best risk adjusted portfolio (looking at the Sortino ratio). The results were

What is interesting is that the computer replicated the barbell strategy (80/70% safe, 20/30% risk) in all portfolios but in one (stock+gold). Stocks had a 4.5% standard deviation (3.51% SD of negative returns only) and gold had a 4.19% SD and 2.43% SD of negative returns. So they are both pretty volatile, which is why the computer didn't find much benefit in barbelling that.

Bonds and bills are far less volatile so the computer quickly found the that it was beneficial to barbell it. This tells me that when mixing assets, it is highly beneficial to think in terms of a 'drawdown' anchor (safe/stable assets) and a 'volatile asset' (usually stocks). That is how one achieves good risk adjusted returns (and a high sortino ratio SHOULD, most of the time, be a good indicator of the best CAGR to max DD portfolio, among other risk adjusted metrics)

I believe that's why the barbell works, it has a very good and stable 'drawdown anchor' (the safe allocation, which is very high at 70-90%) and a small volatile asset allocation. When these two get unbalanced, that's when the drawdowns occur (when stocks do their thing and drop a lot, as they like to do) and risk adjusted returns suffer

This is not to say that bonds are perfect, in fact, I was perhaps too hard on cash in the past, I can see now that the true value of cash lies in being a 'drawdown anchor', limiting total drawdowns. in that respect it can serve a very useful function. Its that smart bond, cash, gold mix that can produce a really powerful drawdown anchor that will be resilient to deflation/inflation/stable periods. with that 'rock' in place, one can then afford to take stock risks (in fact, a lot of stock risk, like EM equities, small cap stocks, levered hedge funds, etc) while still having small drawdowns due the presence of the anchor. i think that's a good way to think about core resilient portfolios

 

[Daal]

A smart cash, bond, gold mix would have to include certain amount of country/currency diversification for tail risk protection.

Buffett loves to say "put all your money in low cost stock index funds" but I believe that is inferior to the alternative (at least for people who know what they are doing). The alternative is to have drawdown anchors in place but take MORE risk in the risk allocation.

Perhaps an aggressive reversion of this would be 60% safety (smart cash,bond, gold mix) and 40% a lot of risk (EM equities, small caps, startups, risky IPOs with stops, levered hedge funds etc). While the 100% index guy will brag about how he is 'beating all the managers out there', the 60/40 guy will almost never lose more than 25% of this money. The index guy will be facing suicidal thoughts as he blows half of his money in 2008, Buffett will be calling everyone in DC to make sure Berkshire doesn't go bust (And Thorp said Buffett thought BRK could go down in 2008), the 60/40 guy will be down only 15-20% while only trailing the index guy by something like 1-3% on average during good years
A better approach IMO

 

[Daal]

The best risk adjusted portfolio on earth is probably something extreme along these lines. Lets say 99% safety, with an expected long-term real return of 1% (developed market bonds with gold in it) and 1% in maximum risk equity (early FB type companies). Most years nothing will happen, the portfolio returns 0-1% (so flattish) but when those companies work out, the returns will be huge (10x-100x)
So the Sortino ratio, CAGR to max DD, Sharpe ratio, etc everything will be off the charts. A lot of these metrics will soar simply by the fact that there is barely any denominator in their formulas (since the portfolio doesn't lose money) but the numerator will be huge
I believe thinking in terms of these shorts of barbells can help a greatl deal when designing portfolios and deciding how to take risk inteligently

 

[Daal]

In fact, the perma 'buy index funds' guys would probably do a much better service to people by saying 'instead of 100% stocks, put 40% in bonds, 10% in gold and 50% in EM equities'. While the barbell will still be unbalanced with the latter (too much risk), its a lot better than 100% risk. There is less tail risk as well, less drawdowns, less overall volatily. So you give up a little in return but you get SO MUCH for it.
A much better risk adjusted return, which to the average person (prone to panicking and making dumb mistakes) will be HUGE. In theory 100% stock index funds is great, in practice (when used by flawled humans) I don't think its so great.
Chasing that last 1-3% in return will lead to stress, panic, anxiety, shorter life spans, heart problems, mistakes, other unknown problems, its very likely NOT worth it for most people

 

[Daal]

So the way to produce great risk-adjusted returns is to have a 'drawdown anchor' in the form of safe assets and take a lot of risk in the non-safe allocation. A decent rule of thumb is 70% safe (diversified currencies, bonds and metals like gold) and 30% risk (high convexity ETFs like EM ETFs, Small caps, Tech, others). Noticed I mentioned high convexity not 'high return', the return depends on a lot of things but the convexity is there. Since high convexity implies a lot of upside, historical returns are not reliable (so even if they say, EM returns less than Developed, I call bs on the study). You put in a few outliers in there and the historical result is completely different. Similarly with small caps, you put a few huge gainers in there and it will beat big cap by a good margin, even if in the last 40 years it didn't do. Key is to put the risk bucket in those 'high convexity plays'.

Can you overdo? Absolutely, 30% as 5% plays in tech startups can lead to a big drawdown quickly as they all go bust. Key is to size according to risk tolerance. High convexity ETFs are great because they are a lower variance version of picking statups. So it will have a higher % of winning years vs those (and some years, will be monster gains vs the lower convexity stock ETFs)

But the beauty of barbell style portfolios is that you have a much better idea of what kind of drawdowns you will face. with 100% index funds the drawdowns can be anywhere from -1% to -100%. With the barbell (with its anchors and the risk allocation being limited) it will be capped at something like 20-40%. Much smarter IMO