Pattern of consecutive lower lows and lower highs, price has broken decisively through support to significant new lows, fundamentals & news flow bad and deteriorating, surprises are negative rather than neutral or positive, declines are larger and longer-lasting than rallies, market sells off hard on bad news, grinds lower on no news, and finds it difficult to sustain rallies on good news, support levels get broken and resistance levels hold firm, etc.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Global Macro Trading Journal
- Thread starter Daal
- Start date
Pattern of consecutive lower lows and lower highs, price has broken decisively through support to significant new lows, fundamentals & news flow bad and deteriorating, surprises are negative rather than neutral or positive, declines are larger and longer-lasting than rallies, market sells off hard on bad news, grinds lower on no news, and finds it difficult to sustain rallies on good news, support levels get broken and resistance levels hold firm, etc.
Quote from Specterx:
Let's back up; I said "I'd like somebody to show me some data proving that the difference between the 10-year bond yield and the SPX dividend yield has historically been a good predictor of stock market returns."
You then appeared to interpret this as a claim that "treasuries at 0.1% a year are just as good an investment as blue chip stocks at 20%." Of course they aren't; but neither is simply observing the difference in yield sufficient to decide what, if anything, you want to buy with your hard-earned cash. If you want an illustration of why, I'll go back and pull some "stocks are cheap" articles from 2000 and 2007 employing similar 'reasoning' to that used by Goldman in its recent call.
Any division between investing and speculating is always artificial; in practice the lines are blurred. Even a 'real-economy' business investment requires, at a bare minimum, making predictions about the future path of input costs and consumer demand patterns. I'm quite comfortable with seamlessly blending the two when I see fit - and quite convinced that those who refuse to do so under any circumstances, and refuse to admit the artificiality of this belief, are missing a piece of the puzzle.
Ok, firstly it's about earnings yield, not dividend yield.
Secondly, it's impossible to use historical data to prove causation in investment, because one cannot conduct controlled experiments. Any one of many variables could have been contributed to the returns, so we cannot isolate - using data analysis alone - which input causes which effect. To do that, one needs to use logical analysis from first principles. Although I would note that when stocks have yielded more than bonds, they have generally outperformed over the long-run (so much so that this yield edge disappeared for decades).
Basic investment theory shows that investment return comes from the cash you can get from an investment. That comes in the form of cash payments from dividends, retained cash that determine net asset value, and the value of future earnings. Since earnings yield (minus essential spending to maintain competitive position) is an indicator of what cash payments you are likely to receive, or see reinvested in the business, then clearly earnings yield matters to investment return.
Yes, there are uncertainties over future earnings, and over the price the market will pay for those earnings at any given time. But no one is claiming that a higher yield differential between stock earnings yield and bond earnings yield is a *guarantee* of superior returns. It is simply being claimed that the greater that differential, the more likely stocks are to be better value and thus provide superior returns, and the more likely bonds are to be poor value and provide mediocre returns. And that statement of probabilities is all you need to justify putting more in stocks and less in bonds - there are never any certainties in investing and it is not necessary to prove something certain in order to support any investment decision.
The division between speculation and investment is as clear as the difference between thin people and fat people; just because there is a point at which they merge and blur does not mean they aren't substantially different concepts. The fact that a 150lbs woman might be on the border of voluptuous vs overweight does not negate the fact that a 90 lbs woman is thin and a 300lbs woman is obese. Speculation is not investment just because the latter has some risks that can't be fully removed.
For example - paying 3 million to take over a company with 10 million cash and no debt, then immediately liquidating it, would be an investment that requires NO forecasting of future business prospects. Buying Apple calls the day before earnings is a pure speculation and nothing to do with investing. Pretty much everyone accepts that those are two meaningfully different concepts.
The distinction is not artificial, but based on meaningful differences. Investments are based on an intrinsic value floor to secure safety of principal, they try to keep risk conservative and they make money from internal returns (cash yield + change in NAV) and any capital gains are from prices recovering to reasonable estimates of intrinsic value; they are bought using long-term money and if debt is used then it is to a level easily financed by sustainable internal cash-flow, and secured by marketable assets. An investment is generally based on analysis of factors that can be understood and promise a high probability of an acceptable or superior return. Speculations generally do not promise safety of principal, high risk is accepted if reward is expected to be high, they often use leverage or short-term funds, they make money from market returns rather than internal returns, and capital gains are essential and based more on changing investor sentiment than intrinsic value. The uncertainty is far higher, and the wager relies on things that are much less amenable to analysis and reliable forecasts. A speculation can involve selling short, using options, making pure bets, with no way to find intrinsic value; and investment cannot. An investor can make money without there even being a stock market or a price quote - just by collecting the yield. A speculator requires someone else to buy at a higher price (or reverse for a short sale), while an investor can just sit and collect his dividends.
How does that imply that it's worthless crap? Come on, if you are going to discuss something, don't waste time by using straw man arguments and trying to derive meaningful conclusions from them.
Firstly, the risk premium there was less than 2%, compared to 6% a few months ago - a gargantuan difference. Why didn't you cite my post made at the exact bottom in August, where I argued that yield differentials signalled stocks were a great investment relative to bonds? Any investment approach, you can cite good calls and bad calls by different people, that doesn't mean the method sucks or is great.
Secondly, their earnings forecasts were catastrophically wrong. Garbage in, garbage out. How does this apply any more to estimating investment returns on stocks, than to estimating investment returns on bonds, or estimating the P&L from speculating/guessing on stock market price movements? In every single case, if your forecasts are crap, your returns will be crap. That says nothing about whether the model, strategy, or the theory is good or not, it simply says that the inputs were worthless.
The only conclusion that is implied from the article you posted is that inputting very poor forecasts to a model dependent on forecasts will get a wildly inaccurate prediction of returns.
Firstly, the risk premium there was less than 2%, compared to 6% a few months ago - a gargantuan difference. Why didn't you cite my post made at the exact bottom in August, where I argued that yield differentials signalled stocks were a great investment relative to bonds? Any investment approach, you can cite good calls and bad calls by different people, that doesn't mean the method sucks or is great.
Secondly, their earnings forecasts were catastrophically wrong. Garbage in, garbage out. How does this apply any more to estimating investment returns on stocks, than to estimating investment returns on bonds, or estimating the P&L from speculating/guessing on stock market price movements? In every single case, if your forecasts are crap, your returns will be crap. That says nothing about whether the model, strategy, or the theory is good or not, it simply says that the inputs were worthless.
The only conclusion that is implied from the article you posted is that inputting very poor forecasts to a model dependent on forecasts will get a wildly inaccurate prediction of returns.
Quote from Specterx:
This is an article from April 2, 2007 titled "Cheapest Stocks in Two Decades Signal Bull Market":
http://www.bloomberg.com/apps/news?pid=newsarchive&sid=aIEj7C7IEUeI
So you see, it's worthless crap. Double whammy since that "6.53%" appears to have been based on forward operating earnings estimates.
Quote from Daal:
As far as I know the Fed model(Comparing stocks yields to USTs bonds) has been shown to fail to predict stock returns even for the longer term. I believe Hussman has some articles on that
Firstly, a necessary condition for something like the Fed Model, or (superior) using the equity risk premium, is that the earnings estimates are reasonably sound.
Second, you don't try to 'forecast' stock market returns, this is impossible. You try to wait for high percentage, high reward/risk setups, where the odds are strongly in your favour. The market is semi-efficient and so the fact that stocks have an earnings yield advantage over bonds is not usually predictive - firstly some of it is compensation for risk, secondly some of it is efficient pricing (e.g. anticipating recession). However, when such valuation measures get to extreme differentials, especially when caused by unusual extremes of emotional decision-making and sentiment amongst investors, then they do become more predictive of the relative odds of outperformance.
Finally, it's easy to disprove this disproof, by using reductio ad absurdum - Hussman would have to claim that if stocks yielded 100% per annum, and USTs yielded 0%, that this would not predict superior returns for stocks. Since that is obviously nonsense, so is his analysis.
The correct inference from that analysis is that the Fed model, or the superior equity risk premium model, does not accurately forecast returns when within normal bounds. However, if you analyse the results when the valuation disparity is at more extreme levels, and do some basic overlay for the robustness of earnings data (i.e. don't rely on street forecasts if there's an obvious credit bubble), then it does become significantly predictive of the odds of future outperformance.
HK elections this weekend. I plan to buy a lot of HKD on monday and hold this large position for 1 month, then go back to the normal position
I believe the chances of a reval will be higher than normal after the transition. I just hope I don't chicken out because if this trade works out it would mean a lot to me in terms of being set for life financially and the risk is not significant(spread and carry is very small for 1 month, fluctuations are limited as the chances of deval have to be so small I would have to live several lifes on avg to get hit by a black swan of this kind)
I believe the chances of a reval will be higher than normal after the transition. I just hope I don't chicken out because if this trade works out it would mean a lot to me in terms of being set for life financially and the risk is not significant(spread and carry is very small for 1 month, fluctuations are limited as the chances of deval have to be so small I would have to live several lifes on avg to get hit by a black swan of this kind)
Quote from Ghost of Cutten:
Yes, there are uncertainties over future earnings, and over the price the market will pay for those earnings at any given time. But no one is claiming that a higher yield differential between stock earnings yield and bond earnings yield is a *guarantee* of superior returns. It is simply being claimed that the greater that differential, the more likely stocks are to be better value and thus provide superior returns, and the more likely bonds are to be poor value and provide mediocre returns. And that statement of probabilities is all you need to justify putting more in stocks and less in bonds - there are never any certainties in investing and it is not necessary to prove something certain in order to support any investment decision.
I did address this in my original post:
It's certainly possible - even likely - that the returns on stocks will be higher over the entire lifetime of the 10-year bond. That is, higher than 2.4%/year or whatever, nominal, not compounded. That is over 10 years from current prices and says absolutely nothing about whether there'll be a 50% decline somewhere along the way, etc. It doesn't say anything about your real returns which could be deeply negative. Not my idea of a generational buying opportunity in stocks.
These are not trivial concerns. On the other hand, it's hardly received investment wisdom to realize that 5% is greater than 2%. If all you're doing is tinkering with percentage weights between stocks and bonds in a bog standard always-in, always-long diversified portfolio, then this sort of analysis is probably sufficient. I just assumed the bar was higher here.
The division between speculation and investment is as clear as the difference between thin people and fat people; just because there is a point at which they merge and blur does not mean they aren't substantially different concepts. The fact that a 150lbs woman might be on the border of voluptuous vs overweight does not negate the fact that a 90 lbs woman is thin and a 300lbs woman is obese. Speculation is not investment just because the latter has some risks that can't be fully removed.
For example - paying 3 million to take over a company with 10 million cash and no debt, then immediately liquidating it, would be an investment that requires NO forecasting of future business prospects...
Your buyout example would almost certainly not occur in reality, given a competitive market etc. And that's exactly the point: in actual practice, any investment worth making (and plenty that aren't worth making) involves material assumptions or forecasts about the future, in other words a degree of speculation. To continue your analogy, just as most people are neither obese nor rail-thin but somewhere in between (normal distribution), so most capital allocation decisions involve elements of both investment and speculation. We employ the separate concepts to organize our thoughts, but there is in practice no sharp impermeable barrier between them. Rather, they represent opposite ends of a continuum.
Quote from Ghost of Cutten:
How does that imply that it's worthless crap? Come on, if you are going to discuss something, don't waste time by using straw man arguments and trying to derive meaningful conclusions from them.
Firstly, the risk premium there was less than 2%, compared to 6% a few months ago - a gargantuan difference. Why didn't you cite my post made at the exact bottom in August, where I argued that yield differentials signalled stocks were a great investment relative to bonds? Any investment approach, you can cite good calls and bad calls by different people, that doesn't mean the method sucks or is great.
Secondly, their earnings forecasts were catastrophically wrong. Garbage in, garbage out. How does this apply any more to estimating investment returns on stocks, than to estimating investment returns on bonds, or estimating the P&L from speculating/guessing on stock market price movements? In every single case, if your forecasts are crap, your returns will be crap. That says nothing about whether the model, strategy, or the theory is good or not, it simply says that the inputs were worthless.
The only conclusion that is implied from the article you posted is that inputting very poor forecasts to a model dependent on forecasts will get a wildly inaccurate prediction of returns.
Well, I was comparing the embarrassing failure of an opinion piece by one Wall Street investment house made several years ago, to a recent opinion piece which appears to be similar in its claims, nature and methodology. A particular market call of yours from last fall is really neither here nor there. Of course as we move forward we'll get more information about how well our respective methods stand up.
There are several reasons why the 2007 example is relevant. You blame incorrect earnings forecasts; has Wall Street since become more accurate or robust in its methods of forecasting earnings?
Has the model been updated or extended to incorporate the criticisms I mentioned in my post? That is, making sweeping claims about "generational buying opportunities" evidently without regard to downside risks, absolute returns, or the secular market cycle?
I would venture the answer is "no."
Quote from Ghost of Cutten:
The correct inference from that analysis is that the Fed model, or the superior equity risk premium model, does not accurately forecast returns when within normal bounds. However, if you analyse the results when the valuation disparity is at more extreme levels, and do some basic overlay for the robustness of earnings data (i.e. don't rely on street forecasts if there's an obvious credit bubble), then it does become significantly predictive of the odds of future outperformance.
Here at last we establish the goalposts and lay out the assumptions. Sounds to me like the basis for a potentially actionable strategy.
Some questions:
1) What do you consider to be 'extreme' levels?
2) What filters or qualifiers are used to interpret available earnings data? A simple discount from analyst estimates (or rather, waiting for 'really extreme' levels) might provide the needed margin of safety.
3) Is there any easy way to convert 'outperformance' into reliable and acceptably large absolute returns e.g. in an environment where future inflation may be unpredictable, or where future absolute returns on both stocks and bonds are expected to be poor?
Quote from Daal:
HK elections this weekend. I plan to buy a lot of HKD on monday and hold this large position for 1 month, then go back to the normal position
I believe the chances of a reval will be higher than normal after the transition. I just hope I don't chicken out because if this trade works out it would mean a lot to me in terms of being set for life financially and the risk is not significant(spread and carry is very small for 1 month, fluctuations are limited as the chances of deval have to be so small I would have to live several lifes on avg to get hit by a black swan of this kind)
What's your rationale behind this trade? My opinion is that the HKD reval (/repeg to the RMB) will be a gradual process. Hong Kong depends alot on its exports after all, so a sudden revaluation (appreciation) would disrupt alot of businesses. Was wondering also what sort of bid/offer you're seeing on the HKD?
Quote from shawnp:
What's your rationale behind this trade? My opinion is that the HKD reval (/repeg to the RMB) will be a gradual process. Hong Kong depends alot on its exports after all, so a sudden revaluation (appreciation) would disrupt alot of businesses. Was wondering also what sort of bid/offer you're seeing on the HKD?
See the 2011 Ackman presentation on the HKD. I agree with the RMB thing, they are not going to change the currency, not they will de-peg but they CAN change the level of the peg against the USD. Henry Tang confirmed that a couple weeks back
The 2014 Fed pledge combined with the latest rally in risk markets, their inflation rate + sky high real-estate + Election makes me think there is a decent chance they will reval. Perhaps 10-20%
I don't recall my bid ask friday, I use Oanda