It's very timely to take another look at stockmarket seasonality. The brief panic last week (barely a 2% selloff) actually came at a moment where seasonality has turned to an increasingly negative bias. The seasonally bearish bias peaks in later September to early October. So it's fair to say that all else equal, the risk of a correction or further selloff(s) is at elevated levels over the coming weeks and months. Having a solid investment process and analytical system (indicators, framework, etc) to know what to do, should a correction come, will be key.

Despite the seasonal bias, corrections typically don't happen spontaneously without some sort of catalyst. But if you think about it there are a few potential candidates out there e.g. North Korea, Trump/US political risk, Iran, Jackson Hole, September Fed meeting, mid-October debt ceiling deadline, and just plain old fashioned market risk e.g. high valuations and increasingly buoyant expectations. So whether it's hedging or low-beta plays, or valuation cushions, taking a cautious tone may be a good idea for now. Makes me think of the value of cash - one asset which arguably becomes undervalued when markets get overvalued.

The 2017 YTD path of the S&P 500 overlayed against the historical seasonal pattern. The purpose of using two axes is to compare the signal rather than the absolute % changes.

On a similar note, here's the CBOE VIX YTD compared to the historical average levels - coincidentally, this is about the time you actually expect to see VIX trending upwards.

For completeness, here's a more timeless chart which shows the seasonality of the S&P 500 and the seasonality of the VIX - they move much as you would expect.