---
title: 'Market-timing model: Do investment banks time the market? (Excel)'
source: 'https://www.youtube.com/watch?v=kPdVYFxg_bU'
video_id: 'kPdVYFxg_bU'
date: 2026-09-18
duration_sec: 1105
channel: 'NEDL'
---

# Market-timing model: Do investment banks time the market? (Excel)

> Source: [Market-timing model: Do investment banks time the market? (Excel)](https://www.youtube.com/watch?v=kPdVYFxg_bU)

## Summary

This video explores how market timing, rather than stock-picking (alpha), explains the performance of major investment banks. Using 10 years of data and a CAPM-based regression model, it demonstrates that JPMorgan's outperformance was driven by accurate market timing, while underperformers like Credit Suisse and Barclays exhibited significantly negative timing coefficients.

### Key Points

- **Intro and Channel Plug** [00:00] — Welcome message, request to subscribe, Patreon support, and acknowledgment of SeedSpot's top 20 list. The video topic: market timing as a source of outperformance.
- **Thesis: Alpha Isn't Everything** [00:46] — Market timing can beat the market even with zero or negative alpha. It's about forecasting market movements and adjusting risk exposure (risky vs. safe assets) accordingly.
- **Sample: 7 Investment Banks** [02:33] — Data on Morgan Stanley, Goldman Sachs, JPMorgan, Citigroup, Credit Suisse, Deutsche Bank, and Barclays over 10 years (Oct 2010–Oct 2020), using S&P 500 as market proxy.
- **Performance Results** [03:44] — All banks except JPMorgan underperformed S&P 500. Deutsche Bank, Credit Suisse, and Barclays had negative annualized returns. JPMorgan was the only beat.
- **Model: Augmented CAPM** [05:15] — Adding a squared market return term to CAPM captures market timing. Coefficient gamma: positive = correct timing (move with market), negative = incorrect timing (contrarian moves hurt).
- **Regression Methodology** [08:40] — Calculated daily returns, squared benchmark returns, ran linear regressions (market return + market return squared) for each bank; reported coefficients, standard errors, and p-values.
- **Morgan Stanley: Positive Timing** [11:36] — Alpha insignificant (no stock-picking edge), beta >1 (risky/leveraged), but market timing coefficient significantly positive, indicating valuable forecasting.
- **JPMorgan: Outperformance via Timing** [13:15] — JPMorgan had near-zero alpha but positive market timing (second-best after Morgan Stanley), explaining its superior stock performance.
- **Negative Alpha Cases** [14:36] — Credit Suisse and Deutsche have significantly negative alphas (poor stock-picking). Credit Suisse and Barclays also have negative timing, worsening outcomes.
- **Graphical Conclusion** [16:23] — Scatterplot shows underperformers (Citigroup, Credit Suisse, Barclays) have poor timing; Deutsche suffers from bad picking; successful banks (JPMorgan, Morgan Stanley) have positive timing.

## Transcript

Hello everyone and welcome again to another the best platform around for distance learning in business, finance, economics and much much more. Please don't forget to subscribe to our channel and click that bell notification button below so that you never miss fresh videos and tutorials you might be interested in.
We would also greatly appreciate if you consider supporting us on Patreon. Please check the link in the description for more details. Finally, we would like to express our gratitude to SeedSpot for including our channel into
their top 20 YouTube channels on economics. Thank you very much for that shout out. But without further ado, my name is Saba and today we are going to discuss market timing, a crucially important yet often overlooked aspect of portfolio management and a quite
promising source of outperformance and abnormal results. that you can successfully use, potentially, to beat the market by exploiting your foresight and superior forecasts.
And traditionally, when we think of funds, investors, portfolios beating the market, we immediately start thinking about alpha. Well, alpha is fine and good.
It's mostly associated with, investment managers speaking, undervalued stocks that deliver positive risk-adjusted returns. But it needn't be the whole story. For example, if you are an active portfolio manager, it's very unlikely that you just pick stocks that you deem undervalued and stick with those.
It's most likely that you actively move your money from asset to asset or even from asset class to asset class. And here, it needn't be the case that you pick undervalued stocks.
It can be the case that you just forecast market movements correctly and invest aggressively into risky assets when you expect the market to go up, and disinvesting from risky assets
and staying in cash or government bonds, risk-free assets, or even hedge assets or safe-haver assets we have discussed in a previous video, when your forecasts tell you that the market is likely to go down.
If you correctly anticipate such market movements and actively move your money accordingly, you can beat the market, extract abnormal returns, even having a zero or negative alpha.
And sometimes this can be even more important than picking undervalued stocks. But how important is it exactly to explain the performance of most prominent hedge funds or investment banks?
Well, today we are looking at exactly that. We have got data on 7 prominent investment banks Morgan Stanley, Goldman Sachs, JPMorgan, Citigroup, Credit Suisse, Deutsche Bank and Barclays.
All of those have at least a listing on New York Stock Exchange. Well, some of those are primarily listed in Europe but at least one listing they have on New York Stock Exchange. And we will consider how important is market timing in terms of explaining the fluctuations
of their corporate value. as those are investment companies, or at least they have significant or substantial investment divisions, their corporate value should reflect the net asset value of their portfolios, at
least closely following it. So, from their stock price fluctuations and their relationship to movements of the market, which is proxied here by S&P 500, as usual, over the period of 10 years, from mid-October
2010 until mid-October 2020, we will be able to extract what is the contribution of market timing and alpha to the performance of these seven investment banks.
Well, first of all, what we can do is we can see how well do they perform over these ten years. To do that, we can divide the total return index at the end of our sample period by the
of return index at the very start of our sample period, raise it to the power one tenth as we have a 10 year period that would give us annualized return and subtract 1. And we can enforce this formula and drag it across our 7 investment banks and the market
and see that, well, some of those performed exceptionally poorly. Primarily Deutsche Bank, well, we all know about the scandals and the issues they had with this punishment lately, but also Crazy Swiss and Barclays also had a negative annualized return.
The only bank that delivered a return in excess of S 500 was actually J Morgan While Morgan family Goldman Sachs and Citigroup delivered positive returns but well below that of
the market. So overall it has been a pretty bleak decade for investment banking. But what allowed JP Morgan to achieve this rare outperformance, at
least surpassing all of its close competitors we can see here? Was it alpha? Was it the fact that J.P. Morgan asset managers were correctly picking $100 stocks? Or was it because J.P. Morgan asset managers could correctly predict where the market is going and move their money accordingly?
G2, asset portfolio management and market timing. Well, that's what we're going to investigate just now. Well, as it is often in finance, a model that can measure some important concept is most
often just a clever generalization of the well-known capital asset pricing model that relates the expected return of a stock or an asset in general to the return of the market.
And obviously, in the general case, those are excess returns or net of the risk-free rate, but as we have daily data, the impact of the risk rate on the returns is negligible, so we can omit it from our analysis.
So we can relate the asset return to the return of the market, to the beta of this particular asset, to its alpha, and the rest, the unexplained volatility, would be captured by the disturbing
term epsilon. Well, here, in this framework, we will be able to measure how well does the asset perform in terms of its alpha, in terms of risk-adjusted performance, and how risky is the asset in
terms of market risk, in terms of systematic risk, in terms of its beta. But by augmenting the Kappern model with market return squared, we would be able to see how well the asset performs, or the portfolio performs, or the investment fund, in that
case an investment bank performs when we consider the impact of market timing. Because if we introduce the market fund squares into our equation, we would be able to see
does the fund do remarkably worse or remarkably better when there is a strong movement in the market in either direction, and that's what the square term captures.
If this coefficient gamma, the market timing coefficient, is positive, it means that the manager of the fund, or asset managers in an investment bank, correctly anticipates stock market movements and invests in risky assets, leverage their balance sheet or leverage
their portfolios when they correctly predict that the market will go up, and disinvest from risky assets when they correctly predict that the market will go down. If it's negative, that's the other way around. It means that you incorrectly predict where the market will go, overexpose yourself during bearish periods, and have some foregone upside during the bullish periods.
Well, that is how one can approach this issue asymmetrically. to measure those coefficients, alpha, beta and gamma, alpha being the measure of stock picking ability, beta being the measure of systematic market risk,
and gamma being the measure of market timing ability of a particular asset manager, we can measure those using linear regressions. But first, we need to calculate the yield returns of our investment banks,
the benchmark and the square benchmark return. So, first of all, calculating the returns. Full return index today divided by the total return index yesterday minus one. Drag that across all of our investment banks plus the benchmark,
bottom right click all the way down, and then calculate the return of our benchmark of our market as in the 500 squared to get our estimations of market timing ability in linear regressions.
And again, bottom right click this formula all the way down across the sample. Now, we have to estimate a linear regression with the return of the market and the return of the market squared for each of our investment banks.
So, we need to select a 2x3 array, because we need, first of all, the coefficients and the standard errors, and we have got three coefficients to estimate, with about two explanatory variables market return and market return squared for BFN market timing respectively and we also need our constant alpha so three columns Now we can just import the linux function and as our y is our dependent variable we can select the
returns of Morgan's family over here and as our independent variables we can select those two arrays q and r with the S&P 500 return and S&P 500 return
squared. We do need our constant because that's our alpha that we care about out quite a lot. So we press 1 over here and we need to report the additional statistics because we want to get the standard errors to determine statistical
significance. So we put 1 here as well, close the parentheses and enforce this formula using shift-control-enter and here we get our regression output. So we can see that first our market-standard coefficient is positive, meaning that in
terms of asset management, active portfolio management and moving money according to your market forecast? Well, Morgan Stanley asset managers are doing quite well. But are they doing significantly well? Well, let's figure it
out by applying the usual T-value. First of all, we need to calculate the key stats by dividing the coefficient by the respective standard error, dragging it across, and then applying the p-value by using the two-tailed t-distribution. And we input
the absolute value of the two stats. And as the number of degrees of freedom, we need to consider that we have got 2,516 observations, but we have reduced the number of degrees of freedom by three, and we have three coefficients to estimate, so minus three. But regardless, if your number of
degrees of freedom is quite large, the selection doesn't really impact the result. So if you have more than 60 degrees of freedom, any additional degree of freedom does not really impact the
significance that much in terms of calculating the p-value from the t-snap. But it's always nice to be extra precise. So we can close the parentheses, can close the formula, and drag it across. And here we see that, first of all, the alpha of Morgan Stanley is negative, but it's insignificant,
meaning that asset managers of Morgan Stanley are not spectacular but not really terrible at stock picking. They are doing stock picking roughly as well as you would have expected them to do if they were doing it randomly.
Obviously, there is a positive beta that is actually quite a lot higher than 1, meaning that Morgan Stanley is either very leveraged in terms of its investment portfolio or is investing in terms of its investment banking activities
into assets that are quite a lot riskier than the market at large, at least judging by the market risk exposure, the systematic risk exposure brought to it by Master Beta. But, most interestingly, market timing is also significantly positive,
meaning that asset managers at Morgan Stanley do accurately forecast where the market will move and withdraw or invest additional funds into risky assets accordingly, meaning that market
farming potentially is the most important feature that contributes to the Morgan Stanley's piracy performance. Now we can do exactly the same procedure for each and every of our investment funds by
copying this formula, pasting it over here for golden sacks, dragging this column with the dependent variable to represent golden sacks returns, again, unfortunately,
using shell.env and reapplying the tset on the p-values. As we have correctly referenced all the cells, if we just copy and paste over here, we can recalculate all of the ttacks correctly. And now we can just proceed with doing
it for every single of our environment banks. Now do JPMorgan and that's something we are actually quite interested about because JPMorgan is the only bank that showed superior performance
in S&P 500 over this 10 year period. Then let's do Citigroup and here we can actually just manually correct the array with the dependence variable from L to M as we know where it's
going to move and here we for the first time see negative market timing meaning it's not universal for investment banks to correctly anticipate where the market will go Now we can consider credit swift and correct M to N again moving the array with the dependent variable one to the right once more and calculating the t on the p and for the final two banks
Deutsche and Barclays. Again, now it's column O. And for Barclays, finally, it's column P.
And what do we see here? Well, first of all, none of the banks that we have considered, at least, have significantly positive alphas. All the more so, none of those have a positive alpha, even if
we talk about insignificant coefficients. However, some of them, namely Credit Suisse and Deutsche, have significantly negative alphas, meaning they have picked assets that have grossly
underperformed the market. Do they provide for it by being better at market time? Well, Deutsche Bank is not market climbing at all. So roughly we could say that as an
investment bank, Deutsche is more passive without moving funds from one equity class to another to gamble on its forecasts of market movements, while two other underperforming banks, Queen Swiss and Barclays, have notably
negative coefficients here for their market climbing and they are remarkably significant both at 1% confidence interval. That means that asset managers at Barclays and Credit Suisse incorrectly banked on their forecast and either have forgone a lot of
upsides from bullish market periods or have overexposed themselves to market risk during the bearish periods, which contributed to their quite poor performance over the past 10 years. If we look at other examples, we can see that JP Morgan, the best-performing
investment bank in our sample, actually also has positive market timing that is surpassed only by that of Morgan Stanley, and also has quite close to zero alpha, with it being negative,
but very low in magnitude, as well as quite low, by the standards of our sample at least, That corresponds to investment effectiveness of this investment bank over the past 10 years.
If we represent our results graphically, plotting alphas and market timing ability on a scatterplot, we can see that there are three banks that perform quite poorly, and those are Citigroup,
Credit Suisse and Barclays, with poor market timing contributing to their demise, at least in terms of the stock price. Deutsche Bank stays as an outlier over here, as we all know, Deutsche Bank has suffered
generally because of poor stock picking, poor asset allocation, and not because of poor market timing, and that is why it stays over here, alone to some extent. And the three banks that are generally the most successful from our sample, especially
J.P. Morgan, are the banks that have an alpha of relatively close to zero, and market timing that is positive and in case of J.P. Morgan and Morgan Stanley, 30% significant.
And that is what has driven the differences in performances of seven most prominent investment banks over the past 10 years. We can see that none of them had positive outputs, but some of those had positive and
significant market time inability of their asset managers, meaning that in investment banking, it can be quite hard to constantly allocate your capital to undervalued assets, but it's reasonably easy, I mean, not for some of the banks, but for some of the banks
it can be relatively easy to correctly anticipate market movements and allocate your capital towards either risky assets or risk-free assets accordingly. And that's all there is for market timing and measuring it using linear regressions
with applications to investment banks. Please leave a like under this video if you found it helpful. In the comments below, I'm eager to see any further suggestions for videos in business, economics or financing you'd like me to record.
And please don't forget to subscribe to our channel. Thank you very much and stay tuned!