Passionate, insightful, contrarian at times and always a true thought-leader in his field, Mark Dowding shares fresh fixed income insights every Friday. His musings on the week cover macro developments, bond market trends and his latest positioning thoughts, with the odd joke thrown in for good measure.

Key Points

• US rates and curve dynamics: Treasury yields remained within their low volatility range this week, while the curve has flattened of late, moving against consensus positioning, in the absence of a new catalyst for further steepening.

• AI disruption debate: fears intensified this week, on the back of a USD2 trillion valuation loss from enterprise software stocks over recent weeks and concerns around AI platforms potentially replacing a significant number of white-collar jobs.

• UK data: the UK labour market has softened, and inflation fell sharply this week. This helped gilts and may pave the way for a rate cut in March, even though the outlook for UK assets remains overshadowed by political risk.

• Japanese inflation trends: next week’s Tokyo CPI points to declining inflation below the BoJ's target, suggesting that market expectations for an April rate hike may be overdone. We believe the next move may come in July at the earliest.

• Macro outlook: markets have shifted from universally bullish AI sentiment to a more nuanced view, with the current infrastructure phase presenting near-term upside risks to growth and inflation before longer-term disinflationary adoption effects potentially emerge.

A downtrade in stocks at the start of the week saw Treasuries extend last week’s gains. However, this price action reversed as the week progressed, and consequently US yields remain comfortably inside the recent, low volatility trading range. Meanwhile, the past few weeks have witnessed some flattening of the US Treasury curve, moving against consensus investor positioning in the absence of a new catalyst to keep driving the curve steeper.

As we have noted previously, a much steeper yield curve is only likely in a scenario where the Fed cuts rates more aggressively than markets currently discount. But with this week’s FOMC minutes raising questions with respect to additional policy easing, we think that market expectations for two rate cuts from the Fed in 2026 will only adjust on the back of more significant revisions to the economic trajectory.

Meanwhile, at a time when broad market levels are not much changed, much of the recent market focus has been between and within equity market and credit sectors. This comes at a time when the debate, with respect to the future of AI, continues to dominate discussion forums.

This debate has intensified over the past several weeks, in the wake of Anthropic CEO, Dario Amodei’s, January essay on ‘The Adolescence of Technology’. The articulation of the threat to civilisation, coming from a ‘country of geniuses in a datacentre’, highlights the potential for widespread economic disruption, with AI platforms potentially replacing large swathes of white-collar jobs, on a 1 to 5-year view.

Since the publication of this essay, and the subsequent release of advances in Claude Code, this has triggered an existential debate with respect to a significant portion of the software sector, where around USD2 trillion has been wiped from the valuation of enterprise software stocks in the past month.

Reacting to this, in his paper ‘The Wrong Apocalypse’, software founder, Andrea Pignataro from ION Group, makes a number of arguments against some of Amodei’s conclusions (with respect to the future for software providers).

However, he nonetheless still outlines a future in which AI platforms quickly move from being fluent in ‘language games’ that handle routine tasks for clients, to a state whereby they can move onto more complex tasks on the back of accumulated knowledge and training, which builds contextual understanding.

At this point, job losses are speculated to accelerate across all areas of professional services. This is seen as impacting demand across the entire economy and even calls into question the future of cities such as London, New York or Singapore, which are heavily reliant on professional service jobs.

Reflecting on this, it strikes us that in the wake of these recent gloomy assessments, which mirror the recent miserable winter weather, there may now be pockets where AI pessimism is starting to be overdone. Yet it is also clear that we are dealing with a very uncertain and challenging investment landscape, at a time when valuations have been relatively extended.

We would also highlight that we currently sit more in the ‘infrastructure’ phase of AI, than the ‘adoption’ phase. This means that, over the course of the months ahead, we will continue to see hyperscalers spending huge amounts in capex, and with a large amount of money being thrown at the economy in a hurry, this could suggest some near-term upside risks to growth and inflation, if supply bottlenecks add to price pressures.

Ultimately, AI adoption could become a major disinflationary force, though it is possible that we won’t see the evidence of this for some time to come.

Given some of this macro uncertainty, we retain a neutral outlook on Treasuries and IG credit spreads. In Europe, we see greater prospects for lower inflation in the near term, though also see bunds rangebound as rising military spend weighs on government budgets and government bond supply. We see the ECB remaining on hold and the identity of Lagarde’s successor as ECB President being a much greater topic of discussion and speculation.

Meanwhile, this week’s UK data may help pave the way for a March rate cut from the Bank of England. A softer UK labour market and moderating inflation also helped gilts this week, even though the outlook for UK assets continues to be somewhat overshadowed by political risk, with Labour looking very likely to suffer a humiliating defeat at next week’s Gorton and Denton by-election.

In Japan, suggested accounting reforms with respect to life insurance companies benefited long-dated JGBs. Low coupon, off-the-run superlong bonds benefitted the most, on the narrative that this would remove selling pressure from this part of the market.

Meanwhile, a declining level for rice prices continues to feed into lower Japanese inflation and it is notable that consensus forecasts for next week’s Tokyo CPI are at 1.4% headline and 1.7% core, on a year-over-year basis. Given that Tokyo CPI releases lead the national print, it will be interesting to see how interest rate expectations react to an outcome that sees inflation below the BoJ’s 2% price target and on a declining trajectory.

In this respect, we think market expectations, which place a 68% probability of a BoJ rate hike as early as April, may appear overdone. We continue to look for two rate hikes in Japan this year, but the next move may not come before July. Moreover, the risk to two rate hikes in 2026 may appear on the downside, should inflation data come in below BoJ forecasts.

Rising tensions between US and Iran saw some regional assets and currencies trade a bit weaker, as speculation of US military intervention continues to mount. Should US diplomatic pressure on Iran fail to deliver the desired outcome, there are a number of options and scenarios that could come into play if Trump loses his patience with Tehran.

Within these scenarios there are also a myriad of outcomes with respect to what happens next, and consequently it becomes very difficult to predict, let alone to position for such an uncertain geopolitical outcome.

Although the US dollar has gained over the week and a risk-off move could benefit the greenback, should there be an extension of downside risk, triggering position capitulation, we are sceptical that the dollar will rally too much on a sustainable basis and consequently short-term strength in the dollar could represent a selling opportunity.

Looking ahead

We remain content to keep risk at modest levels for the time being. At a high level, markets appear to be trading in a sideways pattern. However, this masks some of the more extreme volatility under the surface.

On this basis, it would not come as a big surprise to see a somewhat more extended risk-off move occur over the next few weeks, before stocks establish a firmer base from which to rebuild upside momentum.

This being the case, we would still rather wait to add on weakness, and in credit, we feel under no pressure to chase after bonds at a time when we know that supply will remain abundant for months to come.

Returning to the topic of AI, it was also interesting to read reports suggesting that Claude was utilised in the planning and preparation of the US raid to seize dictator Maduro at the start of this year. If only it could be utilised to craft a peace deal that Russia and Ukraine could agree on….but sadly peace in the Donbas remains as elusive as ever, as we move into the 5th year of the current conflict.

Moreover, it is also a concern how AI could be used in nefarious or destructive ways in the years to come. Perhaps it seems certain that we will come to learn that, as with all technology, this can be used as both a force for good and also a force for evil. Also reflecting on this, in Amodei’s description of the ‘Black Seas of Infinity’, he ponders some of the indirect effects resulting from a century’s worth of scientific and economic change being compressed into a single decade. A cheery read, this is not.

In this respect, it may be striking that some of the great AI thinkers of the day appear to paint a picture that we face a nihilistic trajectory towards a dystopian godless future, where humanity lacks any purpose. But such prophecies of doom are certainly overstated.

Moreover, these ideas might also speak too much of the mindset of tech nerds who spend too much time behind their screens and in realms of fantasy and science fiction, rather than in the real world.

Pressing pause on AI is not the answer. We will need to learn to embrace and adapt to change, and thankfully humans are very good at doing this. However, from a financial markets’ context, one clear takeaway from this debate is that the mood surrounding AI was universally bullish for growth and stocks during 2025. In the current landscape, things are a lot more nuanced and challenging.