"If you're playing a poker game and you look around the table and can't tell who the sucker is, it's you." - Paul Newman

Looking at the tentative rebound in gold prices thanks to the fall of the US dollar on conjunction with some respite for some selected Emerging Markets (EM) as well as the escalating tensions between Italy and the European Commission on migrants with pressure building up on the ECB's "Le Chiffre" aka Mario Draghi, when it came to selecting our title analogy, we decided to go yet again for a poker reference. As we stated in our previous conversation, it remains to be seen who ultimately is going to be the ongoing buyer of Italian government bonds with plans for a reduction in QE in Europe.

In the game of poker, tilt is a term used for the state of mental or emotional confusion or frustration in which a player adopts a less than optimal strategy usually resulting in the player becoming over-aggressive (such as Turkey's Erdogan calls on Turks to ignore forex developments and to sell their gold and dollars to support their crumbling lira). Placing an opponent on tilt or dealing with being on tilt oneself is an important aspect of poker.

It is a relatively frequent occurrence due to frustration, animosity against other players, or simply bad luck. Experienced players recommend learning to recognize that one is experiencing tilt and avoid allowing it to influence one's play. In our October 2015 conversation, we praised both Le Chiffre and Mario Draghi. Both of them share the same trait, both are poker prodigies hence our title analogy at the time. We wonder if President Donald Trump could be one as well given the pressure he has applied in his various negotiations on numerous issues including trade tariffs.

In this current game of high stake poker, particularly with China, one might wonder if Trump is not trying to upset the mental equilibrium of his adversaries, which is essential for optimal poker judgment (NAFTA being the most recent example). Though many advanced players (after logging thousands of table-hours) claim to have outgrown poker "tilt" and frustration, many other poker professionals admit it is still a "leak" in their game but we ramble again.

In this week's conversation, we would like to look at the potential for tactical rebound for EM, as we move towards the important US midterm elections.

Synopsis:

  • Macro and Credit - Could we see a tactical bounce in EM?
  • Final chart - US credit growth still on cruise control
  • Macro and Credit - Could we see a tactical bounce in EM?


Whereas the heat was on during the summer for various Emerging Markets thanks to murderous rampage of "Mack the Knife" (King Dollar + positive real US interest rates), we recently touted MDGA aka Make Duration Great Again. Recently, not only have we seen some small yield compression in the long end of the US yield curve, but as well the dollar has slowed down and fallen a tad ensuring a small bounce in gold prices as well.

Also, credit markets in particular US Investment Grade credit have had more favorable returns in recent months. One can therefore wonder, with the latest NAFTA discussions, if indeed some sort of tactical rally for EM is in the cards on the road to the important midterm US elections.

On this subject, we read with interest Bank of America Merrill Lynch's take in their Global Liquid Markets Weekly note from the 27th of August entitled "EM climbs out of Jackson hole":

"US midterms a winner for EM

After Jackson Hole the focus this week shifts to EU and US inflation, but the real story for the months ahead will more likely be the US midterms. We think they are about to turn from a bearish to a bullish theme for EM through positive news on trade and a peak in the USD. So we expect a post-summer EM rally - though a tactical one before global risk assets, this time including the US, come under pressure in 2019.

Midterms trigger post-summer rally

Since our year ahead report ("Taxing EM") we have been looking at EM through the lens of the midterms - now about to turn from a negative to a positive factor. The fiscal stimulus package and trade tensions have been closely linked to the approaching elections and have weighed on EM so far. Trade wars have become by far the most feared risk, according to our monthly investor survey. Now, however, David Woo expects an imminent NAFTA deal and some progress with China, as the polls suggest that the US administration has a strong incentive to present good news on trade. Importantly, we expect the next round of China tariffs scheduled for Sep 6 to lapse as they would be too painful for the US consumer and the agricultural heartland (via China retaliation).

A glass half full - just about

Predictably, EM sold off during our usual August break - as in 8 of the last 10 ones: the 10-year average EMFX return in August is the 2nd worst after November. However, the volatility occurred on low volumes, and neither external nor local debt saw fund outflows over the summer.

The current level of our EMFX carry sentiment indicator implied positive returns in EM carry in 56% of the subsequent 4-week periods since 2004: ie, a glass half full but hardly convincing enough for a big "Back to School" party.

Softer USD = EM outperformance

Time and again a weaker USD dominates any other EM driver, whether global or local. Chart 2 shows this on the basis of monthly correlations of EM returns with the USD and other potential drivers.

This has also been the case during periods of Fed tightening - as long as other factors have managed to hold the USD down. In fact, monthly EM returns have surprisingly been negatively correlated with the G3 central bank balance sheet.

Indeed, our G10 team now sees the USD rally this year close to its peak. (1) Trade war fears seem to have become USD positive lately, so their easing should be USD negative. (2) The US-EU data surprise differential seems to have peaked too for now. (3) Positioning in US vs rest-of-the-world assets is back to historical highs. Short term, Italy budget fears could still weigh on EUR/USD, but for next year our team targets 1.20.

QT doesn't matter … until US HY cracks

Quantitative tightening per se isn't EM bearish: it still comes down to its impact on the USD. If the USD really weakens into 2019, a repeat of the experience of the last cycle is likely: US high yield peaked in mid-2007, then it moved sideways for one year while EM debt continued to perform. However, when US credit finally entered an outright bear market mid-2008, EM wasn't spared either, and risk-off became USD bullish (Chart 2). A key difference to '07 is that EM is unlikely to benefit from rating upgrades (Chart 3).

Consensus says EM fundamentals are still decent: we don't disagree on this specifically, but we do think that the rate of change of these fundamentals is negative, risks are to the downside and spreads are fair rather than cheap even after the latest sell-off.

Best local value: Indo, Mex, SoAf, Thai

Summing up, we see a post-summer rally in EM that benefits local more than external debt as its main trigger would be a weaker USD and less trade fears: it may last for a while, but by 2019 we expect EM debt to crack when US high yield does. To position for a tactical rally in September in Asia we like long THB FX and Indonesia local 10yr bonds. We have turned from bearish to neutral CNY, and local front-end bonds are starting to look attractive after the rebound higher in yields in the past two weeks.

In LatAm we like short EUR/MXN on a better global backdrop and NAFTA. In EEMEA South Africa has the best risk/reward in FX and rates given its high beta to China and EUR/USD. Valuations are good based on the real yield; three hikes priced in; and ZAR near our fair value estimate. EUR/CZK has downside if EUR/USD goes higher.

In contrast we don't find risk/reward compelling in Russia and Turkey. In Russia we have argued for a while that sovereign sanctions risks were underestimated. We just closed our shorts but want to wait for a sell-off on an actual passing of the sanctions bill pre-midterms before buying ahead of the likely oil-bullish Iran sanctions in November. In Turkey real yields are finally in line with the peers, even accounting for the imminent CPI spike, but they are not high enough to compensate for the political noise (Chart 4).

Long term, sovereign external debt looks cheap with an implied CCC rating, but short-term investors have to brace for further volatility spikes driven by the exchange rate." - source Bank of America Merrill Lynch

Flow-wise, this summer has been a bloodbath for many asset classes with no one being spared. These funds' outflows have continued recently according to Bank of America Merrill Lynch, which is somewhat revealing as well the magnet appeal of the US dollar as per their Follow The Flow Report from the 24th of August entitled "IG, HY, Equities and EM debt all down for the year":

"Outflows continue from IG, HY, govies, EM and equities

It seems that investors have nowhere to hide. Almost all the risk assets we follow recorded outflows last week. We saw outflows from IG, HY, Govies and EM debt; same in equities. Higher risk assets' volatility, EM FX sell offs, trade wars and Italian political risks have instigated a risk off trend in flows across risk assets.

Rate differentials across the globe are shifting flows away from low yielding and risky markets. Note the significant shift of assets out of euro credit to dollar credit over the past months on the back of dollar strength.

Over the past week…

High grade funds recorded another outflow last week; but arguably more moderate to that of last week's. High yield funds were hit again by outflows. Looking into the domicile breakdown, European-focused funds recorded the lion's share of outflows, while outflows from global and US-focused funds were more moderate.

Government bond funds recorded another outflow last week, albeit more moderate to the outflow suffered a week ago. All in all, Fixed Income funds recorded another sizable outflow last week.

European equity funds continued to record outflows for the 24th consecutive week. $56bn has left the asset class over that period.

Chart 1: IG, HY, Equities and EM debt all down YTD (cumulative flows, $mn)

Global EM debt funds recorded another outflow last week, while the magnitude of the outflow more than tripled w-o-w. Commodity funds recorded an outflow.

On the duration front, reach for "safety" prevailed. Small inflows into IG short-term funds, were more than offset by outflows in the belly and the back-end of the curve. Mid-term funds suffered the most last week, with outflows increasing steadily over the course of the past three weeks." - source Bank of America Merrill Lynch

The US has attracted inflows particularly given cash is becoming more and more attractive on a relative basis and for investors seeking some sort of safe haven in the context of EM gyrations. What is interesting to note as of late is the clear deceleration in some macro outputs in the US such as Existing home sales declining 0.7% to 5.34 million unit pace in July, new home sales falling as well thanks to the issue of "affordability" in conjunction with the softness in Durable goods orders slipping 1.7% in July. This is in contrast to the latest FOMC minutes showing Fed members clearly in the strong economic outlook camp. This is of course ensuring another hike in September.

Yet, some positioning in US bonds remains elevated we think, and if indeed some pundits get poker-tilted, then there is a probability of seeing some sizable strong short covering in very short order. Same apply to some shorts in the gold mining space, which have been relentlessly hit by the fall in gold prices but it seems we are not there yet. Returning on the "contentious" issue argued by many on the stretched positioning on US Treasury notes, we read with interest Bank of America Merrill Lynch's take in their Global Liquid Markets Weekly note from the 27th of August entitled "EM climbs out of Jackson hole":

"Extreme positioning elevates risk of bond rally

While we continue to target 10y rates at 3.25% for the end of the year, the ongoing buildup in non-commercial short positioning in the Treasury futures complex does not bode well for our outlook. We studied the connection between positioning - as defined by the Commitments of Traders report published by the Commodity Futures Trading Commission - and subsequent changes in 10y rates and found a relatively high occurrence of 10y rallies after non-commercial positioning hit extreme short levels.

While positioning is just one piece of the duration puzzle, the ongoing trade negotiations and increased political headline risk moving into the midterm elections are also not favorable to our bearish rates outlook. On the flip side, however, we still view the minimal Fed pricing - just 70bp of hikes priced between now and the end of 2019 - as perhaps the most compelling reason to remain in the bearish camp for rates and in the flattening camp for the curve.

Chart 5 shows the net position of traders classified as non-commercial, or speculative, across the curve.

This classification is the residual of the commercial designation, which the CFTC defines as traders who use futures to hedge business risks. We also looked at the newer CFTC designations: leveraged, asset managers, dealers, and other, but found the strongest results within the original data set.

The net spec positions shown in Chart 5 are defined as spec longs minus spec shorts, where the units are 1000s of contracts. This is a measure of positioning within the spec category, and by definition will be offset by positioning in the commercial category (Chart 6).

As a result, a study of non-commercial positioning should show mirror image results in the commercial data, which is what we found.

To each week of reported data we attach a simple z-score of the net position by looking at the current value versus the average over the past 2.5 years. We looked at both raw net positioning and net positioning relative to total open interest for the contract. The results were very similar. The data is reported on Fridays and corresponds to the Tuesday of that week. We tracked the 3-month change in 10y rates starting on the Tuesday of each week.

Go commercial

To summarize: when positioning becomes highly unbalanced between commercial and non-commercial traders, ie when a large net spec position or equivalently a large net commercial position develops, the market tends to move in favor of the commercial positions and against the non-commercials over the following 3 months. Table 1 summarizes our results for TY contracts. The results for US contracts were similar.

There are 3 key takeaways:

  1. When positioning becomes unbalanced, there is a higher occurrence of the 10y rate moving unfavorably for the spec positions. This is based on a comparison of how frequently 10y rates rise or fall over a 3-month horizon in the entire sample versus how frequently rates rise or fall when accounting for positioning. As an example, when the z-score of net positioning is lower than -1 (positioning is below its average by at least one standard deviation), we found the 10y rates rallied 63% of the time versus a full-sample rally frequency of 52%. Similarly, when looking at the net commercial position, we found that a net long position beyond 1 standard deviation resulted in a rally frequency of 60%.
  2. The magnitude of 10y rate moves following position buildups increases as the position size becomes more extreme. For example, when short spec positioning is below average (z < 0), the 3-month rate rally averaged 6bp. When z < -1, the average rally was 11bp, and when z < -1.5, the average rally was 14bp. A similar type of tiering was seen for bond market selloffs following buildups of long spec positions.
  3. Positioning typically gets partially covered in the 3 months following an extreme buildup. When non-commercial positions get short and commercials get long, the data show that on average, these positions partially unwind over the next 3 months. This position squaring shows a greater magnitude as position imbalances increase. For example, when net specs measure z < 0, the following 3-month cover is on average 28k contracts. But when spec shorts are z < -1, the 3-month average covering move is closer to +50k contracts. This provides evidence that the rate moves are exacerbated by position squaring, both on the spec side and the commercial side.

The latest data point was observed on 17 August for positioning as of 14 August and we will get an update this afternoon for positioning as of 21 August. The latest z-score for TY specs was -2.2, which occurred in only 4% of the days we looked at back to early 2003. The one bright side for bond bears, however, is that 10y rates have already rallied about 8bp since 14 August. But if the z-score remains high, it would not increase confidence in our 3.25% target for year-end.

Source: Bank of America Merrill Lynch

One would therefore wonder if it's time to put back on your MDGA (Make Duration Great Again) hat on given there is a higher probability that the speculators at the table might at some point in the near future get poker tilted and start short covering massively should the macro picture deteriorates sensibly over the course of the coming months due to internal or geopolitical factors coming into play (the known unknowns).

The key to putting on your MDGA hat will be the direction of the US dollar in the coming weeks/months. Right now, it's still "risk-on" and we continue to favor a US overweight over the rest of the world, yet there could be some relief rally continuing for some oversold EM countries in the mix.

Returning to our poker analogy, us being fond of game theory, we read with interest Bank of America Merrill Lynch's Global Economic Weekly note from the 24th of August entitled "What good is a poker face if you aren't playing poker?":

"What good is a poker face if you aren't playing poker? Last week we discussed the wide array of risks that the markets will have to navigate in the coming months. This week we focus on a common theme in what we view as the three headline risks on the horizon: the US-China trade war, the Iran oil sanctions and the risks emanating from Europe (Brexit and Italexit). These are all conflicts that live at the intersection of economics and politics. Political economists and other commentators have sometimes depicted such conflicts as games of poker or "chicken," in which both sides use confrontational rhetoric to signal a position of strength, in the hope that the other side will capitulate. Here we argue:

  • The "game of chicken" view of political conflict assumes a zero-sum outcome: serious accidents are avoided.
  • But aggressive rhetoric increases the political cost of backing down. Mutual escalation might be the only feasible outcome.
  • In other words there is a risk of stumbling into a "prisoner's dilemma," in which both sides incur serious economic costs.

The devil is in the details

The "game of chicken" and the "prisoner's dilemma" are two popular game-theoretic frameworks that are often used to describe real-world situations. They are similar in many ways. Using terminology from our previous work, in both settings, two countries (or any two sides in a conflict) must choose whether to escalate or de-escalate a potential conflict. Mutual de-escalation preserves the status quo but mutual escalation results in a "lose-lose" scenario that leaves both sides worse off than they were under the status quo. And in both settings, a country "wins" when it escalates the conflict but the other country de-escalates.

There is only one real difference between a game of chicken and a prisoner's dilemma, but it has important consequences. The framework represented in Chart 1 becomes a game of chicken when the unknown variable c, which is the cost of capitulation (losing when the other side wins), is not very large (c<10). In this case, if one country is certain to escalate the conflict, the other prefers to de-escalate. That is, countries would rather “hand the other side a win” than pay the price of mutual escalation. As a result the game of chicken typically has a relatively benign outcome or “Nash equilibrium.” One country wins and the other loses, but barring some miscalculation, the mutually-painful scenario is avoided.

By contrast, Chart 1 represents a prisoner's dilemma if the cost of "letting" the other side win is so high (c>10) that both sides prefer the lose-lose outcome instead. Since they both have incentive to escalate the conflict regardless of whether the other side escalates or de-escalates, the only possible result is mutual escalation.

No one wants to be chicken

We think this simple framework provides broad lessons for many of the political risks facing the global economy. The first is about rhetoric. In a game of chicken, strong rhetoric is helpful in signaling commitment to a confrontational stance. Examples include:

Examples include:

  • The Trump administration's threat to put tariffs on another $200bn in imports from China and raise the tariff rate to 25% (from 10%), even though the prices of consumer goods such as electronics will likely increase.
  • China's unwillingness to compromise thus far, even though economic activity appears to have slowed slightly because of the trade war, and the Shanghai Composite Index is down almost 20% year-to-date. China has responded to the latest US threats by promising to retaliate with tariffs on $60bn of US goods.
  • The strong statements that have been made in the US-Iran conflict. For example, President Rouhani recently said that "war with Iran is the mother of all wars," even though military conflict is not a realistic possibility.
  • The UK government's claims soon after the Brexit vote that it had a strong negotiating hand. More recently, Brexiters' insistence on a clean break from the EU, even though a 'hard' Brexit would be economically very damaging for the UK.
  • The euroskeptic Italian government's proposals for aggressive fiscal expansion, even though it would run counter to EU rules, which call for a reduction in the structural deficit. Chiara Angeloni and Gilles Moec expect the 2019 Italian budget to target a deficit of less than 2% of GDP. This would be modest relative to campaign promises but would still be about double the previous government's 2019 target.

In each case we think the goal is to make the other side blink in the face of aggression. However, the risk is that these conflicts will morph into prisoner's dilemmas. How so?

First, very strong rhetoric increases both sides' political cost from backing down. In our view, China would be willing to make some compromises to the US, including increasing its purchases of US agricultural and energy products, and taking stronger measures to protect the intellectual property of US firms operating in China. And in a game of chicken, given China's exposure to US demand, such compromises would be in China's economic interests.

But it is difficult for China to make concessions when the US is dialing up the rhetoric. The Trump Administration's chief economic adviser Larry Kudlow recently stated on TV that China is a "lousy investment" and is "in a weak economic position," which is "not a good place for them to be vis-à-vis the trade negotiations." This sort of language makes compromise tantamount to an admission of weakness. Little wonder then that President Xi's confrontational stance against the US remains popular in China.

Similarly, it is probably not in the US' economic interests to enforce its sanctions on Iranian oil to the strictest degree possible. As we argued last week, driving supply out of Iran to zero would likely push oil prices above $100/barrel, which would be a headwind to growth. But the threats out of Iran raise the political costs of reconciliation and make the US-Iran conflict more like a prisoner's dilemma. A mutually-and in this case globally-painful outcome could be in the offing.

The second, related concern is that strong rhetoric increases one's own costs from backing down. Over the course of the Brexit negotiations, it has become increasingly clear that the EU will drive a hard bargain because conserving the single market is more valuable than undermining it to keep the UK in the EU. Moreover, the EU committed itself to a tough stance by publishing the EU Council's negotiation mandate before the start of the negotiations. The mandate is all but impossible to change because all 27 countries and the European Parliament would have to unanimously agree to do so.

Some UK leaders, including Prime Minister May, seem to have realized that the UK is in a position where it must compromise, not least because of the looming March 29, 2019 deadline. This may be why they presented a Brexit deal last month that was economically negative for the UK-free trade in goods, the EU's comparative advantage, but not in services, the UK's comparative advantage-but was still economically better than no deal.

However, the proposal sparked rebellion among the Brexiters in her party, with threats to bring down the government. May's "Chequers deal" proposal now appears to be in serious trouble. Brexiters have made the overly-ambitious promise to leave the EU but somehow continue to trade frictionlessly with the region. Now their political credibility concerns may be raising the risk of a no-deal Brexit.

Is talk cheap?

It might be possible to win a game of chicken by "bluffing," i.e., convincing the other side that you are likely to escalate the conflict even if that is not the case. This is not an ideal outcome for the other side but at least there is room for compromise while avoiding an accident. In a prisoner's dilemma, however, there is no value in bluffing or posturing because the other side will always call your bluff.

The conflict between the Italian government and the EU is in its early stages. It is possible that Italy will be able to elicit some compromise out of Brussels on budget targets and immigration by essentially bluffing on its willingness to leave the Eurozone. But to achieve this outcome, Italian leaders will have to reduce the political costs to the EU from making concessions. There should be room to sell any deal as a compromise rather than a repudiation of the EU.

Similar lessons can be applied to the trade war, Iran and Brexit. In each case there is room for compromise and the question is how much each side will give up. To a degree, strong statements can help because they might push the other side to make more concessions. But if the rhetoric becomes too strong, the only deal available is no deal at all." - source Bank of America Merrill Lynch

The questions remain whether the Trump administration will need to tilting or not with their current skills. The jury is still out there when it comes to trade negotiations taking place. Overall markets remain highly supportive, particularly in the US even if some weaknesses have shown up as of late in some macro data, although the Fed is tightening, lending surveys still point out towards economic expansion, even if we are late in this credit cycle.

  • Final chart - US credit growth still on cruise control

The most predictive variable for default rates remains credit availability. The latest Senior Loan Officer Opinion Surveys (SLOOs) point that credit remains plentiful still. SLOOS report does a much better job of estimating defaults when they are being driven by a systemic factor, such as a turn in business cycle or an all-encompassing macro event. Our final chart comes from Deutsche Bank Strategy Update note from the 28th of August entitled "The bearish bias returns" and shows that the latest Fed SLOOs remain supportive of domestic demand (yes US consumer are using their credit cards...):

"Data supportive of higher rates, political risk remains a headwind

Since early August, the data has remained supportive of higher rates. In the US, the bank lending survey has improved and is consistent with domestic demand remaining comfortably above potential (left graph below). GDP trackers for Q3 are consistent with ~3% growth (right graph below), which is broadly in line with our economists' forecast.

It is worth noting that capex has upside relative to a series of indicators (graph below).

This could compensate a negative impact from trade wars on investment. On the inflation front, the latest data was generally encouraging. Our economists continue to expect core PCE and core CPI at 2.3% and 2.5% at the end of next year, which is broadly consistent with leading indicators such as the ISM or the NY Fed Inflation Gauge (graph below).

The recent USD strength could be a concern for the medium term inflation outlook. However, (a) it has been relatively modest so far and (b) given the lead/lags it should become a headwind in H2-19 at the earliest." - source Deutsche Bank

If indeed over the coming months inflation does indeed accelerates thanks to transfer of rising costs from the producers to the consumers, then, we think Jerome Powell the other poker player at the Fed will not be bluffing when it comes to hiking more aggressively and this might still catch quite a few pundits off guard.

"When people use the word 'science,' it's often a tell, like in poker, that you're bluffing." - Peter Thiel

Stay tuned !