Morgan Stanley Thinks We May Have Hit "Peak Fear" And This Is Its Best Trade If We Haven't

Morgan Stanley's trading desk thinks we may have finally seen peak fear. 

According to Morgan Stanley trader Kunal Sodha, "when you have Fed board members commenting on a viral Substack post, it makes me ponder – are we close to a local peak in the fear of AI disruption?" 

As we have done on countless occasions in recent weeks, Kunal notes that despite the S&P cash level close to unchanged since end of October, over the last 4 months there have been notable corrections under the surface of the Equity market (and not to mention crypto where there is another full-blown winter), as index dispersion hits never before seen levels.

Among the more turbulent scares are Growth vs Value, and the Beta L/S factor basket which have corrected >20%, and all of this without a negative macro shock, just repeated AI disruption "scares" which have only grown in size and frequency, culminating with the second biggest crash in IBM stock this century outside of the dot com crash.  

Meanwhile, even as financial conditions have become easier (with the Fed starting QE Lite and on pace for more cuts), hedge fund flows YTD have also become extreme in the context of buying AI beneficiaries and Semis, vs the selling of Software that has become somewhat indiscriminate. As shown below, Goldman's Software vs Semi basket just hit a new record low...

... while Kunal writes that Morgan Stanley's Infrastructure Software basket (MSXXINSW) is now the most sold thematic YTD.

The Morgan Stanley trader then asks the obvious follow up question: what if this is not the peak in fear of AI disruption?

Well, the other side - or hedge, if you will - to all this is an acronym that Morgan Stanley has "borrowed" from first JPMorgan and now, Goldman, namely ‘HALO’ (Heavy Asset, Low Obsolescence) trade, and according to MS, positioning in MSXXHALO is not stretched. In fact, HF Net exposure here is only 53%ile since 2020, and according to the MS trader, clients can buy MSXXHALO 110% June calls for 2.5% here (20.1v, 29 delta). 

Some more details:

As of last night’s close, the S&P cash index was unchanged since end of October but during this journey, there has been significant correction and extreme rotation under the surface:

• Growth vs Value (MSZZGRVL) has suffered a 24% drawdown
• Beta L/S pair (MSZZBETA) has suffered a 21% drawdown 
• S&P 500 Info Tech and Cons Disc sectors have suffered 11% drawdowns

However the index level has been in a chop, because on the flip side 

• Cyc > Def +10%, Industrials +13%, Cons Staples +16%, Materials +22% and Energy is up 25% 

Another curiosity: when looking back at Quarterly returns of S&P Sectors since 2021, this current Quarter is seeing the 3rd largest quarterly return spread between the winning S&P 500 sector and the losing S&P 500 sector since 2021, preceded only by Q2 2025 and Q1 2022. 

And while implied vol (aka VIX) remains relatively subdued, as a consequence of all this chop, 10d realized vol of factors already spiked in early Feb to highest since Tariff sell-off in April 2025, and prior to that Nov-2022 but is now settling back down.

The good news is that the biggest fear coming into this year was that elevated gross exposure and higher factor vol could lead to more disorderly price action and hence a flush lower overall in risk, as the market catches down to the various disrupted sectors;  but headline equities continue to manage through the chop so far: 

How unusual was this last 4 month period been? According to Morgan Stanley, since 2021 when looking at rolling windows of 20% declines in the bank's Beta basket (MSZZBETA), SPX during those periods has been down -11.2% on average, whilst the NDX has been down -16% on average. But one feature of these large Beta drawdowns since 2021 over equivalent time periods has been that macro financial conditions were tighter - not looser as we have seen here: since end of Oct rates are lower, and the $ is weaker. 

From a flows perspective, Morgan Stanley notes that YTD has been extreme per Prime Broker content, which again gives food for thought on heightened trend reversion risk:

• Hedge Funds have net bought Semis and the full AI stack from Power to AI beneficiaries which has pushed a lot of these groups to the peak in exposure since 2020
• On the flip side, it’s been a combination of selling lower quality areas like Unprofitable Tech and then members of the bank's AI Disrupted basket and Software where positioning now sits in the 0%ile since 2020
• As the MS PB desk puts it, “Flows have reinforced the positioning extremes”

So in the context of hits extreme positioning and flows setup, today’s Anthropic Enterprise Agents livestream event, which as Goldman and Deutsche Bank said was a catalyst for the market's rebound this morning, was certainly less bad than feared as it was more geared to enterprise adoption than disruption, and it may have set the market bottom for the time being as Morgan Stanley now sees a tentative bid emerging. As a reminder, this is what Goldman said:

The Anthropic "Enterprise Agents Briefing" kicked off 9:30am, and had contributed to some of the weakness across AI-at-risk complex yesterday (in addition to the already much talked about Citrini blog post). The first hour of trading has been a COVER the news reaction with software complex popping higher.  

Several significant single name moves on Anthropic mentions...many of which were primary constituents of the AI-At-Risk Thematic: CRM, INTU, FDS, LZ, DOCU all rallying ~5%. Market is viewing as more of a PARTNERSHIP w incumbent SaaS vendors vs. DISPLACEMENT. We are seeing HF covers. No significant L/O buyers in the space just yet. 

Deutsche Bank was even more bullish:

After watching Anthropic's Enterprise Agents briefing event, we have even greater conviction that model providers are unlikely to displace software incumbents and are instead positioning themselves and their agents to be an orchestration layer on top of existing and incumbent systems. Various partnership announcements earlier in the day (including Intuit) along with mentioning the importance of data and context that live in the various systems of record and engagement suggest that it remains very difficult to replicate or displace much of the knowledge, metadata, and workflows incumbent systems have amassed ("Claude is only as useful as the data it connects to"). 

This was the market reaction:

Lastly, and going back to how to trade if this is not yet peak fear, Morgan Stanley doubles down writing that the sectors that are benefiting from the AI disruption fear sit within the HALO – ‘Hard Asset Limited Obsolescence’ thematic (which every bank is now pretending is their own, see JPM and Goldman). For those who are worried the disruption has more to go then this group has (much) more to run: 

• The Morgan Stanley HALO basket (MSXXHALO) is built 7 structural pillars: materials, utilities, railroads, pipelines, waste operators, defense, and towers
• In the last year this group is up 28% vs AI disrupted basket down 43% and positioning is not stretched Net Exposure 53%ile since 2020
• 6m 110% calls on the MSXXHALO cost indicatively 2.5% (20.1v, 29dl)

More in the full Morgan Stanley desk note available to pro subs.