After refusing to even consider the possibility of a recession in the US for over a year, the first cracks in Goldman Sachs' armor are starting to appear. Over the weekend chief equity strategist David Kostin said that while the probability of a recession remains low according to GS economists, saying that "their model suggests the US has an 18% probability of recession during the next year and 24% likelihood during the next two years", Goldman's clients and investors "continue to inquire about the impact a contraction would have on the US equity market."

Displeased with this inquiry into the worst case scenario, one which nobody at Goldman predicted as recently as a few months ago, Kostin is forced to present Goldman's sensitivity "of our existing earnings and valuation forecasts to a US recession scenario." Here is Goldman's take of what may be the worst case scenario for stocks in a recession:

We estimate the S&P 500 index will generate $117 in EPS in 2016, a rise of 11% from last year, although EPS growth excluding the Energy sector will equal just 5%. Our baseline earnings forecast assumes the US economy expands by an average of 2.0% in 2016 while world ex-US growth averages 3.6%. We forecast sales, margins, and earnings for each sector of the S&P 500 based on assumptions for US GDP growth, global GDP growth, inflation, interest rates, crude oil, and the US Dollar. 

From a sensitivity perspective, we estimate a 100 bp shift in US GDP growth will affect our S&P 500 EPS forecast by roughly $5 per share. For example, holding other macro variables constant, if US GDP growth averages +1.0% then EPS would equal approximately $111.

And here the humor begins:

However, if the US falls into recession and GDP actually declines by 1.0% (300 bp below our baseline forecast) then EPS would equal just $100 per share, a decline of 6% from last year and a 13% drop from the recent EPS peak of $115 in 2014. Coincidentally, the median peak-to-trough fall in EPS around the 13 recessions over the past 80 years equals 12%. Most investors find the sensitivity of our EPS forecast to shifts in GDP growth much smaller than they expect.

Perhaps it has something to do with Goldman's absolutely abysmal forecasting track record - as we reported before, 5 of Goldman's top 6 trades for 2016 - all of which were bullish - were closed out at a loss just 6 weeks into 2016. Maybe "investors" are finally warming up to Goldman's "predictive" powers.

That however doesn't faze Kostin who says that "simply put, many sectors have low economic sensitivity. The profits for Telecom, Utilities, Health Care, and Consumer Staples firms are relatively immune to swings in GDP."

Right, and just exclude energy, materials, and everything else that is nearing the biggest default wave since 2008, then just apply Tesla's Non-GAAP adjustment, and all shall be well.

It gets better: apparently not only are investors overly worried about GDP growth, they are also panicking about EPS:

As is the case for changes in GDP growth, the EPS sensitivity to interest rates is much smaller than most investors expect. While the sensitivity of profits of Financials stocks to interest rates is high, with each 100 bp change in bond yield affecting the sector’s EPS by roughly 4%, the impact on EPS of other sectors offsets about 60% of the Financials’ change (see Exhibit 3). The sensitivity of our 2016 S&P 500 EPS forecast to changes in crude oil price has a similar offsetting impact between profits of the Energy sector and earnings of the other nine sectors. We estimate that every $10 shift in the average price per barrel of crude oil has a 29% change in Energy EPS but just a 0.8% or $1.00 per share impact on overall S&P 500 EPS. Our baseline assumption is that Brent averages $44 per barrel (18%below last year). If oil averages $34 per barrel, our EPS estimate would be $116.

We are #timestamping that and we promise to revisit in one year what EPS will be if (and when) oil averages $34 (or less).

And while we are amused by Goldman's relentless optimism according to which even a recession will barely impact the US economy, we were truly amazed by the following:

If the US falls into recession and GDP actually declines by 1.0% (300 bp below our baseline forecast) then EPS would equal just $100 per share, a decline of 6% from last year and a 13% drop from the recent EPS peak of $115 in 2014. Coincidentally, the median peak-to-trough fall in EPS around the 13 recessions over the past 80 years equals 12%.

Sure, if one assumes that the recession will be just an "average" one, then Goldman is spot on. What however Goldman ignores is that every single recession since 1981 has been the result of the Great Moderation's kicking the can even farther with monetary policy, resulting in declines as follows: 11%, 21%, 32% and 57%. The trend here is obvious, and for an obvious reason: each prior recession was the bursting of a Fed-induced bubble, one which was planted to supplant the prior burst bubble, and as a result each subsequent recession was more violent and more acute than the prior.

As a result, we wonder if instead of an average recession, the next one won't be the most violent one yet, one in which the faith of central banks themselves is wiped out, and leading to a near complete decimation in profits. As such, the question for the chart below is: the median line or the progression arrow?

And then this:

Once again: average or progression, because if extrapolating the trend from the past 4 recessions is any indication, then the 2016/2017 recession will see a collapse of roughly 75% peak to trough, putting the S&P somewhere just around 500.

Finally, it wouldn't be Goldman if the bank wasn't desperate to unload some terribly underperforming product to go alongside its rose-colored glasses sunshine call, and sure enough it is.

Stocks with the highest combined price sensitivity to the US economy and the S&P 500 have sharply underperformed the broad market YTD. Our “Dual Beta” basket has dropped by 17.1% YTD versus 5.9% for the S&P 500, a gap of 1,130 bp in just seven weeks (Bloomberg ticker: GSTHBETA). The plunge in share prices of the 49-constituents in the sector-neutral basket reflects investor fears of a recession and negative EPS revisions greater than the market. The basket is specifically constructed to have high sensitivity to changes in economic growth and now trades at a P/E discount of 2 multiple points (14x vs. 16x for the median S&P 500 stock). The typical stock has 26% upside to the GS analyst price target compared with 9% upside to our year-end S&P 500 target of 2100. Investors who subscribe to our view that a US recession in the near-term is unlikely should focus on this basket from a macro and micro perspective (see Exhibit 5 for list of constituents).

And here, dear investors who subscribe to Goldman's view that "a recession in the near term is unlikely", is what Goldman is desperate to sell to you. In fact, Goldman's prop desk has a whole lot of this GSTHBETA to sell to any remaining clients, which of course assumes that Goldman still has clients who are still alive after the "Top 6 trades of 2016" fiasco.

We, for one, would take the other side, just as we would take the other side of that other collapsing product which Goldman is so desperate to offload to muppets, one which as we reported last week it is even making into an ETF to assist its hedge funds' offloading it to clueless retail investors: the GS VIP basket.

We wonder which of Goldman's prime brokerage clients are so desperate to unwind their top hedge fund positions to anyone, that Goldman was forced to make this index into an ETF to "make it easy" for retail investors to load up.