Next week, Janet Yellen has a tough decision to make. If the “diminutive” chairwoman dares to raise rates by a “massive” 25bps, she risks tightening into a tightening, so to speak. China’s bungled attempt to find an elusive middle ground between a free floating currency and a centrally managed currency has created a veritable nightmare scenario where market expectations for further devaluation are forcing the PBoC to stabilize the yuan by burning through FX reserves which, all things equal, should be generally expected to tighten global liquidity and put upward pressure on UST yields. Meanwhile, slumping commodity prices held down by a global economy that’s careening headlong into the deflationary doldrums have served to put enormous pressure on commodity currencies and that, combined with depressed Chinese demand and (ironically) long-running speculation about an imminent Fed hike, have emerging markets on the precipice of a meltdown, a decisively undesirable eventuality that will be virtually assured if the Fed hikes. On the other hand, if the Fed doesn’t hike, it loses any last shred of credibility it had, and indeed, to the extent the conundrum described above was in part created by the Fed missing its window to normalize policy, failing to hike next week will only make the situation that much worse as emerging markets will be even more confused as to when or even if the Fed will ever actually rip the band-aid off, and that confusion and uncertainty could make the situation worse even if the initial knee-jerk reaction to a hold would be a relief rally. And then there’s the ever-present issue of just what the market will read into the decision. That is, a hold risks conveying the Fed’s concerns about the nascent “recovery” which, as we’ve seen, certainly isn’t up to par from an inflation expectations perspective even if fabricated BLS and BEA data paint a pretty picture from a jobs and output perspective, respectively. So what - and this of course is all anyone will care about next week - is a Fed chair to do, you ask? According to Deutsche Bank, the most likely scenario is a “hawkish hold”, in which Yellen will telegraph the FOMC’s readiness to start the normalization process while admitting (if only tacitly) that circumstances have recently conspired to take September off the table. Here’s more: The most likely alternative is the dirty relent or hawkish hold, whereby the Fed passes in September but October or December remain on the table. One potential signal of a hawkish hold is an announcement that the Fed will hold press conferences at all off-cycle meetings – this would significantly raise risks for an October hike. Additionally, a hawkish hold is less likely to generate dissents than a clean relent because the former leaves October and December as “live” meetings The dirty relent should keep flattening pressure on the curve with risk off dynamics – lower equities and commodities. Then there’s the “clean relent” or, as we would characterize it, the “admission of being screwed” scenario in which Yellen holds, at which point nervous, erratic markets will be forced to determine whether Yellen is i) worried about the economy, which could trigger curve-flattening risk-off behavior or ii) confident in the data but just waiting on inflation expectations to stop doing the exact opposite of what they’re supposed to be doing, in which case everyone cheers more can-kicking and an equity rally ensues: The Fed [could adopt] a clean relent or dovish hold, in which the [FOMC] makes an explicit commitment to when they will hike, presumably in 2016, and presumably because of increased focus on inflation ex-shelter and international conditions. Even a clean relent will require careful communication - a clean relent could communicate concerns over the health of the expansion, or it could indicate a Fed that is reasonably confident on the data but is waiting for inflation dynamics to improve. The former would put equities at risk and would likely produce bullish flattening. The latter case is clearly a steepener, and would be bullish for equities. The Fed will be trying to fall somewhere in between such that rates remain stable and equities are stable to higher. A clean relent is likely to produce dissents. Lacker is a likely candidate for a dissent, but if, for example, Lockhart were to dissent as well the implication would be that Yellen and the doves are dominating – and dividing – the committee. Of course the Fed could also say to hell with it, throw caution to the wind, hike, and see what happens although, as Deutsche points out, expecting the market to take solace in the fact that any hike would likely be accompanied by a dovish bias is likely wishful thinking at best: Finally the Fed could raise rates. We remain skeptical that the market would quickly digest a “dovish hike”, whereby the FOMC raises rates but changes the dots to illustrate a lower rate trajectory, and would argue the market is more, not less, likely to increase pricing for December given a September hike. This would remove the anchor to short rates and opens scope for a rapid move to 1.25% 2s with the 2s5s curve flattening, equities lower, and short-dated implied volatility much higher and feeding back into cheaper risk. And then there's China, whose adjustment the Fed could, theoretically try to wait out in an effort to avoid exacerbating the dynamics associated with managing the new currency regime. That, Deustche Bank says, isn't likely a good idea because as we've been at pains to explain since the yuan deval and actually since the petrodollar began to die, this is all just a global linked liquidity regime meaning this is all inextricably connected: We have our doubts that the Fed can “wait out” China and hike once the currency and equity markets stabilize. This is because China’s currency adjustment and equity performance are in part functions of the Fed – a hike would put further upward pressure on the dollar and could cause an intensification of capital flight, additional intervention, and as a result a more rapid reduction in global liquidity. So ultimately, the punchline is that in reality, the Fed can really never hike. That is, at this point, hiking will always end up being a "policy error" where "error" means making a move that reverberates exponentially through centrally planned and increasingly interdependent and correlated markets. But while no time is a "good" time to make a mistake, some times are worse than others and as Deutsche Bank concludes, a September hike would be an example of really, really bad timing: There might never be a good time to hike, but there are definitely bad times, and this is one. September month-end marks not only quarter end, but the last reference date in determining capital charges for global systemically important banks from wholesale funding exposures. Liquidity is poor and unlikely to improve until the beginning of Q4. Liquidity is an important additional impediment to a hike, particularly because markets settled last week with the implied probability of a September move at just under 30%. We think that the taper tantrum of 2013 is a reasonable point of reference for how financial conditions might tighten given a hawkish surprise – they could get much worse. Our simulations of risk-parity strategies suggest rebalancing away from equities and commodities, due to both increased volatility in these asset classes and elevated correlation between them. The implication is that risk- parity strategies could amplify underperformance of these assets given a shock from the Fed, China, or both. Finally, if you read all of the above carefully and came away suspecting that if the Fed does hike, everything might selloff at once, you'd be correct, which brings us to Deutsche Bank's rather disquieting conclusion, namely that there will be nowhere to hide in a market where everything has become correlated and thanks to the growing role of risk-parity in perpetuating selloffs, that correlation could send funds fleeing to the only thing not moving in the wrong direction, namely the dollar and that, in turn, would only serve to exacerbate the deletrious effects of Yellen's "error": A “policy error” rate hike might well result in positive correlations among equities, commodities and bonds, due to a combination of risk off and higher rates. In this case it is not entirely clear how risk-parity funds would rebalance: A potential candidate for inflows would be currencies, and in particular the dollar, which could be the only game in town. Of course, this would only put additional upward pressure on the dollar, reinforcing the “policy error” nature of the hike via additional traded goods price deflation (including commodities), weakness in net exports, and exacerbating pressure on dollar peggers.