Midnight rally

Discussion in 'Trading' started by Rickshaw Man, Mar 24, 2005.

  1. noddyboy

    noddyboy

    Are you long for the weekend?
     
    #1171     Apr 9, 2021
  2. Just put this position on, been at the range most of the day. Sundays index futures session into Monday is just about always a gap up on the open. The odds are in your favor going long overnight. Use risk management and you will make money. BUY THE CASH CLOSE SELL THE CASH OPEN....SOOOOO SIMPLE.

    Late Friday position over week end.jpg
     
    #1172     Apr 9, 2021
  3. Retief

    Retief

    May the odds forever be in your favor.
     
    #1173     Apr 10, 2021
  4. Again, let's look at the actual numbers and not just some fantasy numbers pulled out of thin air. I use the cash close at 4:00 PM as my closing value and 09:30 AM as my opening value.

    1. For 2021, the actual percentage of gaps up is 57 %. The average gap is + 3,25, so there is a positive skew.

    2. Assuming you're actually filled at the open/close and ignoring transaction costs, the cumulative tally of buying the close and selling the open is + 220,75.

    Maximum drawdown is 15,25 points which happened early in the year. Note: The drawdown have been severe in prior perioods.

    That's over 67 trading days so far this year.

    3. So, how does this compare to simple buy and hold?

    So far this year and using Friday's Close @ 4119,25 and the Open the 4th of January @ 3758,50 - the result is 360,75 points, i.e., outperforming the Midnight Rally by 140 points.

    Accounting for transaction costs; it will be even more significant. And since the Midnight Rally is an overnight strategy, one can't assume extra leverage being employed compared with buy and hold.

    TL;DR - The Midnight Rally is profitable so far this year with low drawdown, but it vastly underperforms simple buy and hold which is to be expected during a trending market which sees gains both overnight and in the cash session (61 % of this year's sessions closed higher than they opened).

    upload_2021-4-10_11-43-57.png
     
    #1174     Apr 10, 2021
    Zwaen, comagnum and ffs1001 like this.
  5. chillibean

    chillibean

    if you wanna swing trade, just get long here. Put a profit target at a point where you would be happy with the money made, and wait for it to hit.
    If it dips hard, put a new long on and get a very cheap price
    Trading the ES doesn't need to be complicated because the market has an inbuilt 'edge' (due to strong upside bias) which means we can all make money if we just keep it simple and buy.
    4400 coming too a store near you soon!
    You won't be seeing much under 3900 for a long time.
     
    #1175     Apr 10, 2021
  6. To expand on the post above and with regards to some comments about how there's only crumbs left for day traders in the RTH session:

    The sum of the RTH ranges so far in 2021 is 3072,50 points. That's considerably more points on the table than both the Midnight Rally and Buy & Hold.
     
    #1176     Apr 11, 2021
  7. I imagine you must been equally confident prior to your 200 + point drawdown the last time you bought the high. :)

    I do agree with the strong upside bias, though. If you were to trade only one direction in US stock indexes, up it is. But it won't be entirely risk-free - at least if you're trading leveraged instruments.
     
    #1177     Apr 11, 2021
  8. chillibean

    chillibean

    all turned out nice in the end :)
     
    #1178     Apr 11, 2021
    Laissez Faire likes this.
  9. Buy the f''king dip have worked remarkably well over the last 10 years, although some of those dips have been considerable. So, make sure to have deep pockets. :)
     
    #1179     Apr 11, 2021
    chillibean likes this.
  10. comagnum

    comagnum

    A comprehensive study by the Federal Reserve done last year on the overnight performance, they call it overnight drift (OD).

    upload_2021-4-11_5-33-22.png

    ES futures avg market performance by time of day 1998 - present
    upload_2021-4-11_5-43-48.png



    My notes from the Federal Reserve study:

    Pre-transaction costs, a trading strategy that goes long the S&P 500 futures between 2:00 a.m. and 3:00 a.m. earns large positive returns equal to 3.7% p.a. with a Sharpe ratio of 1.14.

    Overnight drift does not represent market inefficiency and instead is a phenomenon that arises due to inventory management by liquidity providers.

    More than $15 billion (2018) traded through the e-mini contract daily during the overnight session.


    Overnight returns are negatively related to the closing order imbalance of the preceding day.
    Evidence of high frequency return predictability, arising when market makers take on large positions at the end of the U.S. trading day, which they then trade away in subsequent periods as new liquidity traders arrive to the market.

    Positive overnight drifts occurs only on days following market sell-offs (negative order imbalances). When order imbalances are in the bottom quartile (most negative order imbalances), subsequent returns during the OD hour average 7% p.a while returns through-out the full European trading session are equal to 13.45% . However, more importantly for an inventory risk explanation, during Asian trading hours we also observe significant and economically large positive returns equal to 9.8%. Conditional on large market sell-offs, we observe two geographically distinct locations of liquidity provision: liquidity providers offload imbalances to Asian traders during Asian hours and later to European traders.


    For the sample period 2009-2019, we obtain intraday quotes for the aggregate S&P 500 futures limit order book and can therefore trace the dynamics of market depth.

    Consistent with the idea that market makers set their price schedules to induce mean reverting inventory dynamics, sorting on day t − 1 closing order imbalance, we show the limit order book is deeper on the ask (bid) side when closing order imbalance was negative (positive).


    Price reversals should be amplified in states of high volatility since in these states risk averse liquidity providers bear larger quantities of risk. We test this by interacting order imbalance with the level of the VIX index prevailing at the close of day t − 1, and find volatility has a strong amplification effect on the relationship between order imbalance and the overnight drift.


    Post-2010, we provide strong evidence of high-frequency return predictability precisely when the Tokyo financial market opens. VIX changes, t − 2 to t − 1 contain strong predictive power for the date t overnight drift.


    We explore whether news released after U.S. cash market close can explain overnight returns. Indeed, a large fraction of U.S. corporate earnings announcements are released after

    U.S. market close, as are non-U.S. macro announcements. Studying this conjecture we examine hour-by-hour returns conditional on announcement dates and fail to detect a relationship between overnight news and the overnight drift.



    We show that dealers off-load their excess exposure at Asian open as would be predicted by standard inventory risk management models.


    Over the last 20 years, ON returns have been large and positive between the hours of 12 a.m. (midnight in New York) and 3 a.m. Thirty minutes prior to the opening of the cash market in the U.S. at 9:30 a.m., equity returns display initially large negative returns which become smaller in magnitude but remain persistently negative until 12 p.m. followed by a flat return profile until 3:00 p.m. which is a large positive return until the closing bell at 4:15 p.m.


    The gross CTC return is 4.5%,which equals the average yearly return on the S&P 500 index cash over this 20 year sample period. However, the majority of this return is generated during the ON session: between 6 p.m. and 8 a.m. equity returns average 3.1% p.a. More striking than this, a significant proportion of this return, averaging 3.6% p.a., occurs in the window between 2 a.m. and 3 a.m, a return sequence we dub the ‘overnight drift’ (OD).


    Between the hours of 8:30 a.m. and 10:00 a.m., we observe a sequence of negative returns averaging −3.9% p.a., and we dub this sequence ‘opening returns’ (OR).

    The OR is always negative but only statistically significant on Thursdays and Fridays.

    Suggestive evidence as to why the OR occurs only on Thursdays and Fridays:

    More U.S. macro announcements released at 8:30 a.m. on Thursdays and Fridays. Generally, we experience large positive returns leading up to announcements.

    We conjecture that (short-lived) price-reversals following the macro announcements partly explain the negative opening returns. Secondly, we do not observe many FOMC announcements on Thursdays and

    Fridays and we also know that returns typically are positive in the hours leading up to FOMC announcements which subsequently do not revert. Thirdly, we observe most negative earnings announcements days are Thursdays and Fridays.


    For the hours {01-02, 02-03}, approximately {13%, 9%} of days produce zero returns computed from quotes. However, even the median quote return for the OD hour is large and positive equal to 0.89 basis points per day.


    In all days of the week, the 2 a.m. - 3 a.m. return is positive and significant at the 1% level, except for Thursdays, which is significant at the 5% level.


    The OD is systematically positive and significant in each day of the week.
     
    Last edited: Apr 11, 2021
    #1180     Apr 11, 2021
    Zwaen, Vtechno and Laissez Faire like this.