The hawkish chatter by Fed officials from a month ago is gone, but the hawkish bias in so-called base or high-powered money rolls on.

For the third straight month, the inflation-adjusted year-over-year change in M0 (as base money is sometimes labeled) contracted, dipping 3.6% in May vs. the year-earlier level. The red ink is one more headwind for the US economy, which has been posting mixed results this year.

History suggests that negative readings for this indicator’s annual comparisons are linked with elevated recession risk. The red ink in this corner by itself doesn’t ensure that the US is headed for a new downturn. But the negative bias in M0 isn’t helping—especially at a time when several other key indicators are wobbly–payrolls and industrial production, for instance.

Note that M0’s annual pace since last December has dipped in every month except for February. The negative bias implies that the Federal Reserve is pressing ahead with plans to squeeze monetary policy, albeit modestly. It’s debatable if that constitutes a regime change in this corner, but it’s no longer reasonable to dismiss the idea out of hand.

To be fair, not every dip to sub-zero terrain for annual changes in real M0 leads to an NBER-defined recession. Nonetheless, a negative trend in the monetary base is a risk factor and at the moment the potential for trouble is increasing via this data.

Can the US economy continue to expand with a negative M0 trend bias? Maybe, although much depends on how the numbers from other corners evolve. Ideally, other indicators will offset weakness in M0 and elsewhere.

The good news is that consumer spending still looks encouraging, if only in relative terms. Retail sales in May, for example, posted a solid 0.5% gain. But there may be cracks forming, or so one can argue via the slow but persistent decline in year-over-year sales, as discussed last week.

Perhaps a clearer view of the near term will emerge, for better or worse, when the Labor Dept. publishes the June employment report early next month. A repeat performance that delivers disappointing monthly growth will weigh heavily on market sentiment. Would that trigger a formal recession signal? Maybe not, in part because year-over-year growth for payrolls is still rising at a healthy if decelerating rate.

Meantime, monetary policy has become a weight for the macro trend. It’s not clear if this weight will be fatal for growth. But it’s one more sign that the US macro trend continues to face challenges in 2016 and so the crowd’s capacity for dismissing disappointment is effectively nil at this point.