When can traders out-economize the economists?

When can traders out-economize the economists?

In both my public comments and investor correspondence, I have been consistently emphasizing the importance of separating expert opinions from layperson’s hunches. Applying this advice to myself, I try to rely on my strategic edge rather on my economic forecasts.

In fact, taking this point even further, I don’t rely on anyone’s economic forecasts.  But it is not because I consider the analysts who produced those views incompetent. Indeed, there are those with intellectual frameworks I highly respect. Still, for every well thought-out argument, there is an opposite one equally well-substantiated.

Current US economy and stock market bulls and bears are a great example; both camps have excellent charts, statistics, and qualitative opinions. When two experts I respect disagree so strongly, I have no choice but to believe the future is uncertain. And forecasts from either side have little value other than establishing a paradigm.

So are there times when traders can see something economists do not? The jury is still out on this one. I’m convinced more often than not that market professionals delude themselves by thinking their hunches regarding a particular unemployment number or inflation statistic have predictive power.

But I believe there is one area in which traders may have an edge: we tend to have a strong sense of momentum. While the analysts measure where things are, traders think in terms of where things are going. Thus, when a business cycle turns, traders may have a stronger strategic commitment to a nascent economic trend.

I will not use the cycle turning of 2007-2008 as an example – it was too unusual. But I remember well 2000-2001. The NASDAQ index started to collapse throughout 2000, while the Fed raised rates 100bp in the first half of the year and kept them flat at 6.5% for the remainder of the year.  Only towards the end of the year in 2000 did the Fed begin to initiate mild talk of a potential economic slowdown.  

NASDAQ vs FED Funds Target Rate, 2000 – 2003


To be sure, the central bank accepted recession being a real risk early in 2001 and started cutting rates aggressively. But not aggressively enough even though we in the trenches were quite sure that the dot.com bust was leading to the inevitable recession and a much deeper fall in interest rates. For the younger market participants, I will remind that the recession was not caused by September 11th. It was a virtual certainty months before. I even suspect that the terrorist attack, in fact, accelerated the recovery by forcing the Fed to ease even faster and by bringing stimulus into the economy.

I remember that sense of confidence: “blah blah blah, whatever the research says,” the rates are going down! Of course, for every time one could have been right this way and made a fortune, there could be several false starts; traders reading too much into a transitory data shift. I cannot prove or disprove the value of this momentum sense. I am just bringing it to your attention and judgment.

Why am I writing about this today? Over the last few weeks, my momentum sense started to tingle. The recent fall in oil prices and core inflation together with other incrementally soft data are beginning to signal the roll-over of the cycle and the bond markets are running with it. Meanwhile, the analysts are still vigorously debating.

Crude, US Surprise Index, Core PCE (y/y), June 2016 to present


To re-emphasize:  my bullish bond position is not contingent on this current perception but on my strategic framework which is independent of forecasts. The cycle does not have to end today, I can wait to make money tomorrow.

But I can’t help having a hunch – value or no value.

Good luck,

Flat Curves and Recessions

All too often I hear sentences like “The bond market and the stock market say opposite things about the economy” or “A flat curve means recession, so if you expect the curve to get flatter that means you are expecting an imminent recession”.

The purposes of this post are to dig a little deeper into the concepts of concurrencies and causality associated with the yield curve and to challenge some of the common assumptions embedded in the statements above.

While the “Freakonomics” crowd loves debunking confusion between correlation and causality, I will instead write in terms of the distinction between “causes” and “indicators” and, furthermore, between “leading” and “concurrent” indicators.

My strategy mostly deals with causes. In my book, I discuss how causal relationships between two market events can be characterized in terms of necessity and concurrency and how to trade based on such characterization.

By way of example, the Fed moving to tighter monetary policy has a clear causality relationship with a stronger dollar and a flatter US yield curve. As always, some would disagree even with this paradigm. But I am comfortable with this causality being if not as “certainty”, but an at least a “certain likelihood”. 

In my book, I have demonstrated how, in 2014, the stronger dollar was a concurrent necessity with respect to rising rates and thus a dominant trade. That meant by the time the Fed tightened, dollar would have to have strengthened but it might have (and did!) strengthened even without the tightening.

There are, however, highly correlated pairs of market events with a more obscure causality relationship. For example, up until 2016, USDJPY traded in high correlation with all risk assets, including not only the Nikkei but also the S&P 500. In my post, The Tale of Three Shorts, I have discussed some of the complexity there and even suggested that strong JPY trade may be a self-defeating chicken (a thesis yet to be verified). To put in simply, the correlation was obvious to me, but the causality wasn’t.

In such situations, I refer to events as “indicators” rather than “causes”. I have several times heard sentences like “weaker USDJPY spells trouble for stocks”. Such statements irked me as logically flawed, but the fallacy is not immediately obvious. Indeed, stocks tend to go down when JPY strengthens, don’t they?

This is where I introduce the notion of the concurrent indicator. If I look at my screen and see USDJPY up on the day and don’t look at anything else – what is my guess about the stock market? Of course, it’s more likely that the stocks are up. But why shouldn’t I look at anything else? I have a full screen of prices available.

My point is that USDJPY doesn’t convey any information about equities that I can’t see by looking at the equity screens. Yes, usually when USDJPY falls, S&P 500 trades lower as well, but if it did not go down there is no obvious ECONOMIC causality saying that it has to catch up. So, as I see it there is little predictive power to gain from divergence in this pair; what has happened, already happened. And with this mindset I was not at all shocked with the divergence of this year.

5Yr Chart of USDJPY (White) vs SPX (Yellow)


But what about the clearly statistically confirmed “leading” indicators? My contention here is that even those may not convey as much useful information as appears.

As you can guess, I am taking a swipe at the maxim that flat yield curves forecast recessions and steep yield curves forecast robust economic growth, which appears to be accepted by market participants almost without question.

The evidence is undoubtedly strong – flat curves indeed preceded all the recent recessions. But my question is, “does the curve convey any information about the possibility of a recession, we don’t have otherwise?”

Imagine that you have jumped from the roof of a high-rise. Falling past the 10th floor is a very reliable indicator of soon hitting the ground, but what new information does it contain?  Given that you jumped off consciously and are familiar with the law of gravity – none.

What if you fell while sleeping? Well, waking up to see the 10th-floor flash by is definitely good info. But returning the analogy to macro-trading, my assumption is that you are not asleep.

Then what if the law of gravity changes? In this case, the 10th floor may not even be a reliable leading indicator anymore.

Over the last few decades, we got used to certain patterns of the business/rates cycle. Events were happening in a certain order and over certain predictable time intervals:

1.       Rapid economic growth

2.       Tight labor market and inflationary pressure

3.       Rising rates

4.       Yield curve flattening

5.       Hiking cycle

6.       Bear stock market

7.       Rates fall

8.       Growth slows down

9.       Easing cycle

10.   Curve steepens

11.   Stock market rebounds

12.   Rapid economic growth

Notice, any of the events in this loop could be used as reliable leading indicators for any subsequent events. Assuming the loop persists.

As I have discussed in Chapter 4 of my book, historical patterns are important to study because they are more likely to repeat than not. But if we make an a priori assumption that this loop will just keep going on without alteration – we almost wouldn’t need to observe anything else – we would already know what happens next.

In that chapter, I gave an example of using interest rate momentum as a predictor of future stock market price action. That is, I showed a decent fit between the two-year backward looking change in the 10-year note yield and the two-year forward-looking change in S&P 500. In the loop above it would mean bullet point 7 (rates fall) is a leading indicator of 11 (stock market rebounds).

Backward-Looking Change in 10yr Yields vs Forward-Looking Change in S&P500 


Why is there good information in this pattern? My claim is that the information is there because there is CAUSALITY. Lower funding rates lead to improved corporate profits. And this simple paradigm lends extra strength to my pattern argument.

When, however, people make the assumption that 4 (yield curve flattening) predicts 8 (economic slowdown), their assumptions are not backed up by direct causality. This does not imply the indicator is wrong, but rather it makes it more fragile.

In the past decades, 4 (yield curve flattening) was typically associated with 5 (hiking cycle). However, as we have approached zero rates and entered the world of QE, the laws of gravity have changed. Curves in the developed world are flattening in the LOW rate environment and they are no longer backed by the causality of the increased funding cost.

The jury is out on whether in the new environment yield curves will reliably forecast the business cycle. My contention is that they are unlikely to carry any information not already familiar to an alert macro player.  For myself, I will stick to indicators resting on the simple and evident logic of causality.

Rising stocks and rising rates

Diversified portfolios perform well during a hiking cycle – a simple and undeniable observation in a recent Fortune article by Joshua Brown @reformedbroker.


The points made in this post elicit my response, as they touch upon the core of my own strategy. I offer not a rebuttal, but rather a discussion of nuance on some of those points.

Point 1. Success of diversified portfolios during tightening cycles over the last forty years. I believe this is mostly a function of the overall secular bull market in stocks AND bonds. Given the short-term negative correlation between equities and fixed income, the diversified portfolio of stock and bonds has been performing extremely well in the idiosyncratic environment of the last few decades.

Point 2. Bonds don’t do too badly during a hiking cycle. Very true. I have observed this historical pattern and it has been at the heart of my bond bullish strategy over the last two years. The fact is, bond investors are usually well compensated for the anticipation of tightening and major surprises tend to arrive on the side of lower rates.

Point 3. Stocks do well during a hiking cycle. Can we agree with that? Here is how S&P500 performed relative to the Fed Funds target during the last two tightening cycles.

As you can see, in both cases the stock market rally continued throughout the cycle and, in one case, even beyond the end of tightening. Both of those rallies, incidentally, were followed by major bear market.

I don’t argue agianst the statement “stocks tend to go up, as the Fed hikes”, but I wouldn’t translate it into the statement “the Fed is about to hike, thus it is good to own stocks”. This would be confusing concurrency with causality. Indeed, the correct causality statement would be “absent a major inflation threat, the Fed continues hiking only for as long as stocks go up”.

Hence, if we have a GIVEN knowledge that the hiking cycle will go on and on, it is reasonable to assume that stocks will be performing. But with this knowledge we can do many other trades, such as short eurodollar futures or five-year notes, or long the dollar.

As we don’t know when the hiking cycle will end, we also don’t know when the bull market in stocks will turn.

Interestingly, this observation doesn’t undermine the credentials of a diversified portfolio. Indeed, this is exactly how it works: the end of hiking and the bond rally cushion us against a possible stock setback, when the economy turns.

Image by Skarphéðinn Þráinsson

If the market is a bubble, why so many stocks look cheap?


“The stocks in my portfolio are precious gems, immune to any problems in the system”. Yes, I know it’s a fallacy.

So, please, help me sort it out. I have been dragged into the raging debate on stock valuations. I have to confess, I feel out of my league on this issue, when engaging with such thorough experts as @Jesse_Livermore (http://www.philosophicaleconomics.com), a moderate bull, or @jessefelder (http://thefelderreport.com), a moderate bear.

When I write about interest rates or currency, I possess (or at least project) much authority and experience. But I know relatively little about corporate valuations. My inclination to stay long equities over the last few years has been a function of my observing historical patterns and trends and my long-term portfolio strategy, but had little to do with whether I thought the overall market was still cheap.

So let me share my confusion. Reading various pieces on the overall US stock market valuations, I find the analysis ranging from fair to very expensive. Yet when I look at individual companies I know, so many of them still look like a bargain.

What DO I know? I mostly follow banks, tech, and social media. I know nothing of consumer, retail, utilities, biotech and so on.

Let’s start with banks. (I source P/Es from Yahoo! Finance).

$JPM  11.24;  $WFC 13.29;   $GS 10.97

Those guys seem to be priced for various degrees of bankageddon. Indeed, they don’t need any growth to justify their valuations, all they have to do is not to have all their earnings taken away by fines.

Established tech.

$IBM 13.49; $INTC 13.02

What is going out style? Computers or computer chips specifically?

All of the above enjoy a healthy growth of EPS via buyback and cheap funding. Some scoff at buybacks, but to me buying your own shares, when they are cheap, and locking the funding seems like a great investment.

Now let’s move to the growth sector.

$AAPL 16.82;  $GOOGL 26.85; $FB 74.65; $TWTR N/A

Are those the culprits of overvaluation?

$AAPL is certainly not as dirt cheap as it was a couple of years ago (similar to $INTC), but still doesn’t look stretched at all, given the power of their brand and continuing revenue growth.

In fact, all four companies above have virtually indestructible brands. There monetization is at different levels of maturity. But it’s hard to imagine any of them to be a terrible long-term gamble.

So where does the overvaluation lie? Is it all the biotech’s fault?