HomeSolutionsInsightsNewsroomPeopleContactLogin

Acclaimed macro analytics
Portfolio Construction & Risk Management
Trading Solutions
Macro Models
Thematic equity baskets

Announcements
Press
Blog

Today? The Mag7 12mth fwd PE trades at a 2.4 FVG discount to Qi model value, the GS most short index at -0.6 sigma, and the Russell 2000 at -0.1 sigma. So cheap to model but the only notable move looks to be in the Mag7 through Qi’s macro lens.

So what to look for now?

Have Model RSqs sharply collapsed indicating regime shock? NO, macro explanatory power has remained high across major assets through this correction. There is no sudden narrative shift other than recognition that growth is now the top driver rather than disinflation.

Have Macro Factor Sensitivities jumped indicating a macro shock? NO, despite the evident stress across markets, the S&P500 macro beta impulse remained surprisingly subdued. No surprise to see factor sensitivities sharply rise to risk aversion measures. However, sensitivity to financial conditions have actually fallen over the last 4wks. With the index falling below its 50d MA but the impulse not jumping in response is notable – perhaps telling us positioning and seasonality is just as much a likely culprit here.

Is model price momentum showing signs of troughing? Closer to the zone. SPY model price has fallen 6% over the last 4wks. Historically, the trough of the 4wk descent has not been worse than a 10% fall over a rolling 4wk period (outside of Covid). See the chart below. That would be at ~500 on SPY – the 200d MA

If this is not sinister, where is there value? Back to Mega-Cap Tech? The Mag7 12mth fwd PE now trades at a 2.4 FVG discount to Qi model fair value – the 12mth fwd PE was ~2 sigma rich in rich in mid-July and now ~ 2 sigma cheap to model value today.

#2. The size of the Russell 2000 move – more idio than macro-driven – and still +0.9 sigma rich

Qi’s MFERM (Macro Factor Equity Risk Model) shows that the size of the Russell 2000 rally we have seen can not be explained by macro. The dip in 10yr yields, and the dollar, alongside the steeper 5s30s was indeed supportive of the index. However, over the same period VIX rose, energy / copper fell and CDX HY spreads are little changed.

This speaks to this move perhaps reflecting more a re-assessment of lop-sided positioning – but this does the raise the ante into earnings season. See the chart below.

In aggregate, the macro backdrop has actually been a drag across SPY, Mag7, Russell 2000 over this period. Risk aversion, followed by growth expectation factors have been drags across the indices – most in the case of the Mag7

The Russell 2000 is currently trading +0.8 sigma rich to Qi model value. Since 9th July, the index has risen 10.5% but Qi model price only 1.7%. In other words, we may be seeing exuberance that soft landing has been accomplished.

#3. KRE – the biggest beneficiary of falling inflation & easier finance conditions – also +1.4 sigma rich, the most expensive US sector ETF

The recent >25% rally in KRE (regional banks) is a very similar story to the Russell 2000 – the primary driver of the rally has been idiosyncratic rather than macro – see the chart below – only ~5% of the move can be explain by macro.

Regional banks are among THE most sensitive sectors to falling inflation expectations and easier financial conditions – as highlighted below falling inflation, tighter credit spreads and lower real rates are the top 3 drivers.

With the domestic focus of the regional banks, Visa results commentary on the US consumer was interesting – they have rarely missed revenue expectations: “In the U.S., while growth in the high-spend consumer segment remained stable compared to prior quarters, we saw a slight moderation in the lower-spend consumer segment”

KRE is now the most expensive US sector ETF, according to Qi’s macro warranted fair value. See the charts below showing the dislocation between spot price and Qi’s model:

#4. Squeeze in short JPY positions has prompted a bullish signal in EURPY

EURJPY macro-warranted model value is flat-lining. This latest sell-off leaves EURJPY 1.6 sigma (5%) cheap. For EURJPY to be in regime & this cheap is particularly rare.

The 1y correlation between spot EURJPY & Qi FVG looks good suggesting the recent pattern has been for the mkt to mean revert back to macro fundamentals.

Two health warnings:

(1) BoJ next week. One of the catalysts for this fx move is speculation of a possible bigger (15bp instead of 10bp) rate hike.

(2) Positioning is heavily skewed. The chart below shows spec net shorts in Yen futures as per the CFTC data. Short Yen has been one of the most popular trades out there. There are a lot of positions to potentially unwind.

#5. Within Tech, US Software attractive relative to the broader market

At the other end of the sector spectrum, the IGV ETF (US Software) is showing signs of relative value. IGV has been underperforming the broader market through July, despite model value moving higher.

The major macro driver of the IGV / SPY ratio is a steeper 5s30s curve, which has been driving the model value higher. Alongside, the pair is beneficiary of an improving growth / inflation trade-off. The regime can be well explained with 85% RSq.

#2. Aggressive Short Squeeze – more signs of over-exuberance?

As discussed in the prior observation, we have seen a squeeze in market laggards – a proxy for this is the squeeze we have seen in the most shorted stocks. The GS most short rolling basket is +1.4 sigma on Qi’s LT model and +1.8 sigma rich on Qi’s ST model.

Again, Qi sees this as opportunity to fade. See the chart below where we overlay Qi’s FVG vs. GS Most Short Rolling basket spot price. Fading the short squeeze also highlights the vulnerability, tactically, of the broader market.

#3. GDX - a play on lower real yields / slowdown but Qi rich here

Gold is enjoying a great July. The popular narrative is Fed rate cuts plus a Trump Presidency offer a potent mix for the precious metal - less competition from bond yields, plus geopolitical uncertainty / trade wars fueling a flight-to-quality bid. That all makes sense, but it’s always reassuring when Qi’s orthogonal relationships vindicate the story of the day. Looking at the Gold Miners (GDX), two of the biggest drivers of the ETF are indeed negative sensitivity to real yields & tracking GDP growth rates.

That helps explain why Qi’s macro-warranted model value has risen nearly 3% in July. But spot GDX has rallied more like 16%. As a result, it screens 1 sigma (7.7%) rich on Qi. And Qi’s FVG has done a good job of capturing local high & lows in the recent trading range.

#4. UK equities: buy-the-rumour-sell-the-fact?

After years of being unloved, UK equities are having a moment. The “UK as a beacon of stability” argument is getting a lot of airplay amidst hopes the UK economy can turn a corner, & the contrast with political turmoil in France. The result has been strong performance, especially for the domestically-focused FTSE 250.

Macro justifies this move. Qi model value has moved 2.6% in the FTSE 250s favour over the last month. There are signs, however, that the initial post-election euphoria may be due a pause given how far we’ve already travelled since the election. The FTSE 100 / FTSE 250 ratio (using the tracking ETFs ISF / MIDD) is now 1.1 sigma (3%) cheap on Qi. That’s the cheapest the FTSE 100 has been in relative value terms in 2024.

This model has only ever been this cheap when in a macro regime 25 times since 2009, i.e. crudely we only see extreme valuations like this twice-a-year. And back-tests reveal using this FVG as a level to favour ISF over MIDD elicits a 72% hit rate & a +0.9% average return.

This is not a fade the UK observation. But it is a warning that the initial euphoria for the FTSE 250 specifically has started to look stretched.

Finally, note Qi’s model shows a weaker Pound hurts the FTSE 250 more than the FTSE 100. Unsurprising given 250 companies are more vulnerable to a weaker currency & imported higher costs. Which makes Qi’s Cable model especially interesting….

#5. Cable – risk reward suggests fade the rally

After the debate & last weekend’s shooting, the market is fixated on trying to identify the best Trump trades. But some trade expressions are more obvious than others. Perhaps the most contested area is the Dollar.

Many of his policies imply a stronger Dollar but Trump is vocal about wanting a weaker Dollar.

From Qi’s perspective there are no standout G10 fx crosses where the Dollar screens as excessively rich right now. But, for the bulls, Cable is interesting.

The two biggest drivers are interest rate differentials & credit spreads. And trends in both have been pushing Qi model value higher recently. Increased hopes of Fed rate cuts versus a dialling back of BoE cuts after sticky service inflation data, have combined with credit spreads near the tights to push macro-warranted model value 1.6% higher in the last month.

But, on current patterns, even with the respective Fed/BoE policy stance & the tightness of credit spreads, cable should sit around 1.2850 according to Qi’s model. That means the recent rally has taken it 1 sigma (1.4%) rich to overall macro conditions. Back-testing this FVG reveals strong results (67% hit rate, +0.5% average return).

Plus, the 1y correlation between spot Cable & Qi’s Fair Value Gap is strong suggesting the recent pattern is for the market to mean revert to macro fundamentals.

#6. Tighter than normal distribution of Qi Equity FVGs - on edge for earnings season

It is striking at the moment how many Qi equity models are hovering around macro-warranted model value. US sectors show only 3 models are more than 1 std dev away (Biotech ETFs IBB & BBH are both > 1 sigma rich; OIH similarly rich). There isn’t a single sector that is more than 0.5 sigma cheap.

To emphasize this point, see the chart below - across 306 global indices, ETFs, sectors the average FVG is zero. 79% of that universe within +/- 0.4 sigma - 0.4 sigma being the 1 standard deviation away from the cross-sectional universe FVG mean today i.e. we have thinner tails compared to a normal distribution. Perhaps a sign of waiting for the cue into earnings season...

Finally, note a strong correlation between spot 5s30s & Qi’s FVG over the past 2 years. This suggests the mean reversion has tended to happen the “right” way, i.e. the market correcting back to macro fundamentals.

#2. US Yield Curve II – the Trump effect

The flexibility of the Qi framework means we can add additional variables into our models to stress test certain hypothesis.

As mentioned above, tax cuts, tariffs & possible threats to Fed independence mean many see a Trump win in November as inflationary & therefore a curve steepening trade.

Therefore we introduced the Real Clear Politics Trump less Biden betting odds as a new factor in our 5s30s US yield curve model. It is the biggest positive driver of our ST model, i.e. a Trump win does indeed imply a steeper yield curve.

Statement of the obvious? Intuitively, yes. But there is clear value in having the ability to quantifiably demonstrate the fact; the ability to measure the size of the political impact; a means to see how important the factor is over time. Custom models like these will be increasingly important as move closer to the election.

#3. EU High Short Interest names are 1 sigma cheap to Qi model value – vulnerable to a squeeze?

Like in the US, economic data surprises in Europe have fallen into negative territory, GDP Nowcasts have faded and policy uncertainty has been rising. Notably, the Euro Stoxx 50 is down over the last 3mths in sharp contrast with the US i.e. sentiment for the region has sharply deteriorated since Q1.

With that said, we note that Qi FVG of the GS EU High Short Interest basket stands at almost -1 sigma. This is at the low end of the FVG range since January 2023. See the first chart below. This is an interesting observation given it may well reflect sentiment for the region i.e. when the high SI basket is deemed cheap to macro, has sentiment become too bearish and a short squeeze is possible, or vice versa?

The following chart compares the Qi FVG for this basket to the broader Euro Stoxx market itself. The last few occasions where the FVG was heading towards -1 sigma, it was better tactically to be buying market dips as opposed to fading rallies.

#4. No room for disappointment this earnings season, especially for mega cap Tech; The most expensive stock in the S&P500 based on Qi’s model – no surprise, Tesla:

We are approaching the 6 week window where we will see 90% of major global companies report, with the last 2 weeks of July being the focal points with around 30% of S&P500 companies reporting each week. Consensus is expecting ~9% y/y growth for the S&P500, which would be the strongest quarter since 4Q 2021. The biggest six tech companies (AMZN, AAPL, GOOGL, META, MSFT, and NVDA) are expected to see earnings grow 30% y/y, compared with 5% for the rest of the market.

6 of the Mag7 stocks are trading above Qi’s macro-warranted fair value, albeit macro explanatory power (Model RSq) averages only 54% across the names. We highlight the table below – Tesla is trading at its 100th 1yr percentile on Qi FVG; Apple at its 73rd percentile; NVDA at its 71st. Indeed, across the whole index, no surprise that Tesla has the highest FVG – the stock has rallied almost 40% in the last 2 weeks. See the over-extension in the second chart below.

While we have no edge on earnings, the above table would imply a lot has been priced in - any disappointments are likely to be heavily punished. NDX is trading 0.6 sigma above Qi model value; SPY is 0.3 sigma above Qi model value.

#5. Nikkei 225 – Overexcitement over the last month relative to growth expectations?

The Nikkei has hit an all time high. Indeed over the last month, the index has been among the strongest major indices globally (in local currency terms). Even in dollar terms, it has been outperforming the S&P500 over the last 3 weeks. Strength over the last month has been led by Topix Banks and Insurers, alongside higher domestic bond yields. Yet over the last month, Nikkei’s Qi model value has seen a sharp drop!

According to Qi’s machine, the largest macro driver of the Nikkei is not the level of JPY or bond yields, but rather global economic growth. See the first chart below.

Our "Economic Growth" driver accounts for nearly 30% of explanatory power in our Nikkei model. Over half the recent fall in macro-warranted model value comes from the downward re-rating in Now-Casting's tracking GDP estimates.

The Nikkei is today trading at +1 sigma above Qi macro-warranted fair value. While Qi’s model value has started to move higher, the dislocation to spot price is large – see the chart below. We wonder if the recent excitement has gotten ahead of itself, if especially investors are concerned on economic growth slowdown into Q3.

#2. AUDCAD FX to catch possible RBA – BoC policy divergence

This week has seen inflation in both Australia & Canda surprise to the upside. On paper that should dent hopes of rate cuts from both the RBA & BoC, but there are nuances.

The RBA told investors they discussed rate hikes at last week’s meeting &, after the CPI beat, markets have discounted around a 40% chance of an increase in the Cash Target Rate in August. The BoC have already delivered one rate cut & markets still perceive them as being in policy easing mode; the question is more whether stubborn inflation delays further cuts. For example, after this week’s inflation print expectations for another 25bp reduction at the July 24th meeting fell almost in half (from ~60% to ~35%) but it is still a rate cut that is being discussed.

On Qi, the Aussie Dollar screens as modestly rich to its G7 peers. Of the Aussie crosses that are in regime, AUDCAD has the smallest Fair Value Gap. Aussie is ‘just’ 0.3 sigma (0.44%) rich to aggregate macro conditions. That suggests AUDCAD could be the pair with the biggest potential upside if you fear the RBA become a hawkish outlier.

That is also borne out when we drill down to inspect AUD crosses sensitivity to both nominal & real interest rate differentials.

• AUDCAD has the biggest negative sensitivity to Inflation Differentials, i.e. amongst all G7 pairs, higher Aussie real yields benefit AUDCAD the most.
• AUDCAD has the biggest positive sensitivity to Rate Differentials, i.e. a hawkish RBA that lifts Aussie nominal yields relative to other G7 yields, & its AUDCAD again that rallies the most.

Two final observations. Macro momentum for AUDCAD is trending firmly higher. Macro-warranted fair value is now 0.9090, the highest since mid-December. Only AUDJPY has seen model value improve more dramatically over Q2. Which makes sense from a RBA-BoJ policy stance perspective, but long AUDJPY positions face the issue of MoF intervention risk.

Aside from the false call in October last year, Qi’s valuation gaps have done a decent job of marking local turning points.

#3. Ahead of the first debate, risky assets are not yet making thematic bets on a Trump presidency; only conclusion for now is that Trump remains a positive for risk:

We have created custom Qi models for major indices, sector and now S&P500 stocks which include the difference in Real Clear Politics Betting Odds for Trump winning vs. Biden winning.

At the S&P500 index level, we have already noted that rising Trump odds are a supportive upside driver of the index. At the sector level, XLK is the standout. It is the most sensitive sector and sensitivity to Trump-Biden betting odds is the #1 +ve driver for the model on our short term horizon.

We subsequently screened at the stock level to compare the performance of the top 30 stocks benefiting most from a Trump win vs. the bottom 30. This revealed the relative performance shown below:

The full list of names is available on request. Interestingly, the longs are Technology, Healthcare and Consumer Discretionary. The shorts are dominated by Energy & Materials.

At first glance, this would seem count-intuitive given Trump’s tariff policies, in particular a 60% tariff on China imports. The tech sector, in particular hardware, has high supplier exposure to China. However, such distinctions do not appear to be made currently.

The only firm conclusion we can make currently based on the current data, is that the market appears comfortable with a Trump presidency. This will almost certainly evolve as we approach the elections and Qi will update its analysis accordingly.

#4. Sector rotation continues under hood – what is still rich? Staples & Biotech

The sector rotation under the hood of the index over the last week has seen Energy and Financials outperforms, Technology underperform (see last week’s MacroVantage where we highlighted the dislocation). However, selectivity if also required to discriminate between defensive sectors.

US Staples today has the highest fair value gap between Qi macro-warranted fair value and the spot price (+0.9 sigma), relative to any other major sector. The divergence between the Qi model price and the spot price over the last 10 days is notable. See the first chart below.

The falling Qi model value has been a function of VIX, rate vol and credit spreads drifting higher and Eurozone confidence and metals drifting lower.

Our back-testing shows a 71% win rate going short Staples at the current level of FVG and holding until withing 0.25 sigma of fair value (since 2010). Similarly, IBB is trading at +1.3 sigma – a 75% win rate going short at these levels.

#5. AT&T

AT&T is now almost 2.5 sigma rich to its Qi macro modelled fair value.

This makes it almost 15% rich and the stock is at its 100th percentile in terms of valuation over the last year.

The AT&T share price is at 52 week highs having almost closed the gap from its drop in Apr 2023

The high dividend yield combined with lower 10y yields has clearly supported the stock.

Further upside is now dependent on continued easing in financial conditions with the stock positively sensitive to:
• Lower real rates
• Less QT
• Higher liquidity

Somewhat anomalously, AT&T also wants a higher gold/silver ratio

However, the last month has revealed greater uneasiness under the hood: Cyclical sectors have underperformed defensives; the equal-weighted S&P 500 actually saw its ATH at the end of Q1; crowded themes have been attacked (power / Utilities, Copper) and previously winning factors have been more range-bound (momentum). Couple that with the surge in meme stock trading that may be suggesting froth.

It strikes us that for the first time this quarter SPY has finally converged to Qi model value – see the chart below. At this point, there is a bigger FVG discount in equal-weighted S&P 500 (-0.6 sigma) than the headline S&P 500 index. Until yesterday’s CPI release, Qi model price momentum had been treading water over the last month.

2. Thankfully the CPI miss should support market breadth with value emerging in cyclicals – this was key because FCIs / disinflation remain the arbiter of stocks

The most important macro upside drivers for QQQ & SPY? Tighter credit spreads, weaker dollar, lower real rates, lower inflation expectations i.e. FCIs remain key for the macro narrative to keep supporting Qi’s model value.

In this regard, the CPI miss was welcomed. As a reminder, the below table shows the relative winners / losers from the dovish CPI print. Winners are both tech orientated but also value / domestic cyclicals – financials and homebuilders. Into the summer, there may well be scope to rotate into these laggards.

XLF relative performance should see respite – Qi’s LT model for XLF vs. SPY RV is high and stable. FVG has hit -2.8 sigma – the widest valuation gap in recent history. Other large valuation gaps are notable in XLI vs. SPY (-2.7 sigma) and IGV vs. SPY (-1.6 sigma)

3. What next for OAT yields?

After hitting 3.32% in the initial aftermath of the shock election announcement, 10y OAT yields have fallen around 15bp. That reflects international rather than domestic events. This week US Treasuries have benefitted from a strong auction, soft CPI & an unchanged Fed. Lower US yields have been the catalyst for the move in global bond markets.

It is noticeable though that Qi’s model shows 10y OAT yields now sit at macro-warranted fair value. The election spike took yields 0.6 sigma (15bp) above our fair value metric; but that move that has now retraced. What next?

Clearly politics is going to be critical in the weeks ahead. But model confidence of 87% says macro is key too. Qi model value is marching higher. Back in mid-May, aggregate macro conditions ‘justified’ yields just under 3.00%. Today, model value is some 20bp higher.

4. French large caps vs midcaps - a more targeted approach to capturing the Gallic risk premium?

We wrote an insight earlier this week - “Pricing the Gallic Risk Premium”. Here we screened Eurozone equities and EUR FX crosses to highlight the vulnerabilities. We noted (1) SX5E sensitivity to our Eurozone confidence measure was close to long term range highs (2) Qi’s model price for Eurozone equities had already been falling (impacted by falling copper prices / waning China optimism) (3) Screening for most sensitive single stocks, the impact is broader than just the banks.

We also note here that the CAC 40 index itself is not the best proxy for domestic activity given its international exposure. The French Midcap index SBF250 should be a better bearish hedge. We created a Qi model for the CAC 40 / SBF 250 relative value pair. The ST Qi model RSq is 84% and FVG is -1.0 sigma. On the LT model, the RSq is lower but the FVG is -0.9 sigma. See the first chart below. The top driver are lower Eurozone confidence and a higher gold / silver ratio. In other words, this RV pair may offer a more discriminate defensive bet than merely shorting French large caps outright. The second chart shows the close relationship of the pair to the OAT / Bund yield spread.

5. Mexico - VaR shock or carry opportunity?

Hello summer carry? The election-driven sell-off has driven the Mexbol index right to the bottom of its Qi fair value gap. Confidence in our Qi model of the index has been stable for the last 12mths. Investors will be hoping that the market reaction acts as a warning shot deterring any unsettling action from the Sheinbaum administration.

EWW, the iShares MSCI Mexico ETF is -2.4sd cheap to Qi model value. Since 2009, it has never been in macro regime (i.e. model confidence > 65%) and this cheap to model value. See the charts below – Extremes in the Qi fair value gap have been a good gauge for price moves in recent years.

Similarly, Qi’s models for USDMXN have a current fair value of ~17 vs. 18.8 spot price.