Global Macro Trading Journal

[Daal]

Gaming and fooling regulators isn't a hard thing to do, smaller companies or small private planes might gamble and try to game them more often (after all, they got less money for compliance). Eventually it blows on their face and the regulators close the loophole but while the loophole/trick is going on, there is a higher CHANCE of a crash. It just doesn't show up in the crash statistics simply because crashes are rare. When it does show up (the crash occurs), its only 1 crash and regulators learn that they need to close that trick. Meanwhile all the risky flights from before are not computed. And the smaller companies/planes are trying to find a new trick again. It also appears that smaller companies have less experienced pilots so it doesn't help

The reason I'm bringing this regulator gaming up is because, its not only a matter of probability of small plane/company crash vs big plane/company crash. Everybody knows it is higher with the smaller ones, that is already taken into account the Sklansky theory by inputting the historical probability of a small plane crash. Its that if they do these sorts of regulator gaming and other tricks, the historical probability of a crash might be understated.

All the fights that didn't end up in crashes are not going into the crash database but while they happened, they had a higher chance of a crash. Eventually the crash happens and regulators put in extra precautions.

If they consistently game the regulators, eventually it will show up in the crash statistics. The problem is that if the dispersion is big enough (because they dont do it all the time, they don't do it consistently, etc), it might take a REALLY long-sample to have that appear in the data. So the historical chance of a small plane crash might consistently be understated. It might take (and I'm just guessing here) 100-200 years to have enough of a sample

 

[Daal]

When it comes to airplane durability (engines, parts etc). Things are more reliable because physics and data coming from physics tend to be more stable. You put in a human component and all of the sudden it takes a lot longer samples to know what is the chance of something. What is the chance of a market crash? It happens so rarely that you need a lot more data to be certain.
Smaller planes/companies might have an 'extra' human element component as compared to the big planes and companies (like the 1 plane company that was being flown by the owner who tried to save some money on fuel and crashed). As a result, the data that comes out of it might be less stable and reliable

 

[Daal]

It would be pretty cool to test my optimal portfolio (the one that I designed at this point but that I will teak/improve in the future) in the Greek 2008-2016 period. That would be a nice way to see how it does under severe conditions. I'm planning to work on that. So far I got this snipet of data

If someone wanted own 10y or 30y Greek bonds and kept their maturity constant (that is, they sold and rebought a new bond at the end of every year to prevent duration/maturities from dropping down), they earned a ~7% CAGR since 2008. Investors almost doubled their money since 2008! Who knew that? If you all you read is bloomberg and ZeroHedge you would think Greek bond holders are broke by now. I think these returns are better than owning EU equities (Havent checked so don't quote me on that)

Of course, that would have required the investor to not be scared out by headlines to stick with the rebalancing plan (in this case I'm calling it rebalancing the act of keeping maturity constant). But that's the history of financial markets, usually it pays to take risk while everyone is scared to death

 

[Daal]

It will be interesting to see how the actual rebalancing does in this case. Because the strategy would have you buying equities on the way down and we know that Greek stocks are down 90-95% from their peaks. I'm sure it wont be pretty but the bond part and the gold part will at least make sure the investor will stay in the game. The stock market part will lead to a severe drawdown but it will be limited given that my optimal portfolio has less than 30% in equities.
It will also raise the point about the importance of global diversification. If that Greek stock part were 15%-20%(with the rest being other countries stock markets) I'm sure the final drawdown would have been quite a bit smaller

 

[Daal]

That said, given that the backtest will stop at 2016 year end, it will be a quite unfair backtest. It essentially will have taken all the stock market drawdown (with all the rebalancing cash from bonds/gold going into the stock market) but not the recovery. Kinda like testing a US portfolio from 1929 to 1933. But for the purposes of looking at the total drawdown (and having an idea of the risk of ruin of the strategy) it should given an idea of what one of the worst case scenarios look like

 

[Daal]

The results of the Greek backtests are amazing. While the country has suffered, perhaps, its worst depression in history, some of the optimal balanced portfolios even MADE money.

First, the "computer portfolio". This is the portfolio the computer told me it was the best for monthly US data from 1926 to 2016. Note that I'm using YEARLY Greek data, I don't think that matters though. These returns bellow are all inflation adjusted (by the Greek CPI). They cover annual real returns from 2007 to 2016, 1 year rebalancing is assumed

A 13% real gain in a 10 year stretch for a real CAGR of 0.3%. This while the unemployment rate soared to 25% and the country fell apart.

Now the modified version of the computer portfolio, the one I tweaked to be able to take advantage of the 30y bond market duration

Not as good as the computer returns but a -3% real return is an ok result. Essentially, the portfolio kept its purchasing power through a huge crisis

What about the Dalio portfolio that he gave to Tony Robbins?

An ok result, -10% real isn't terrible given the magnitude of the crisis.

Keep in mind the backtest was very unfair, it kept rebalancing gains from bonds/gold down to a plunging stock market all the way down(The stock market total return EUR index fell 90% in nominal terms in that period). The stock market did had a bounce in 2012 and 2013 (+38% and +36% real), which the rebalancing strategy promply 'locked' gains by selling some of the exposure as it went higher. This helped to control some of the damage
But still, it was a very unfair backtest because at some point in the next 5-10 years Greek stocks will have a monster rally and that's when the beauty of optimal/balanced portfolios in depressions will show up (especially mine and Dalio's portfolios, since we rely more on equities)

 

[Daal]

Just to given an idea of the effect of the rebalancing into a plunging stock market, if you remove rebalancing, the real total returns range from 29% (computer) to 18% (Mine) to 12.84%(Dalio). So for the optimal portfolios to produce a net real gain/losses of 13% -3% and -10.89% despite that fact. I consider a huge victory for balanced portfolio investing.

Risk party takes a lot of flack in the media, yet anyone using an approach like that would have protected their wealth in a massive way. And I didnt even add global diversification or a higher allocation to gold (which I'm considering doing in my own optimal portfolio for a number of reasons). With those things in a Greek investor would have done pretty well, despite buying stocks all the way down!

 

[Daal]

"But what if Greece had defaulted?", since these portfolios rely quite a bit in bonds, one might think that they would have lost huge had a default occured. It's possible that this would have happened but its also possible that they would have made even MORE had Greek defaulted

I consider the EUR a 'gold standard', it prevents countries from easing monetary policy. If Greece had defaulted, presumably (and the ECB said that I think) they would have to leave the EUR. The new drachma would plunge huge against the EUR. This would be similar to the countries that left the gold standard in the 30s. The effect of leaving gold in the 30's was a huge 'reflation boom'. Stocks soared, the price level went positive (from a deflation), gold soared.

So its possible that the stock portfolio would have soared (after a fair amount of rebalancing buys). The gold part certaintly would have soared (even in real terms due big inflation and real asset chasing) and the bond part certaintly would have a recovery, maybe 50% or more. Plus, over the years, growth would return (helping stocks) and bonds would continue to pay their (new) coupon. Stock booms after depressions end tend to go a lot longer than people think.
Overall, Greece and, I believe, its investors would be better off
So I don't think a default would be all that bad for these portfolios

 

[Daal]

The real returns that were assumed in the backtest

 

[justrading]

Daal, I saw a Tweet that on the White House website one of the changes is they aim to cut taxes, including corporate taxes.

How Trump will fund infrastructure with reduced revenue and not blow up the deficit is a wonder to me. House Republicans are not going to accept a ballooning deficit, which means something has to give.

That will affect market expectations and sentiment.