WhoTrades Live provides thousands of portfolios that you can view, follow, copy, and learn from more broadly. It allows you to invest on your own without being alone in the process and without the large costs associated with outsourcing managing your savings to expensive financial advisors and actively managed products. Recent returns, drawdowns, trade activity, and current holdings are all transparently displayed so you know exactly what type of portfolio, trading style, and risk/reward characteristics you’re following or investing in. This if the fifth and final part of this series, in which I covered different asset allocation approaches. Part I had to do with building a portfolio to meet minimum returns expectations. Part II was about the expected returns versus risks of various asset classes and leveraging them to return the same, what the resultant risk amounted to, and how to combine these by weighting them in such a way to reduce risk relative to a more traditional portfolio. Part III discussed a second approach that involves using the concepts discussed in parts I and II to design a portfolio that takes a “market neutral” stance. But this done by picking individual securities or other traders’ portfolios that you like rather than simply indexing. Part IV covered a third approach involving investing into the assets that you like without the restrictions of the templates that made up the first and second approaches. Part V shows semi-empirically how having multiple diversified revenue streams improves your reward per each unit of risk. In Part I, there are diagrams showing this effect at various correlation levels using matrix multiplication to plot out the relationship. ___ There will be three asset classes used – US stocks, US intermediate-duration bonds, and gold. The US went off the gold standard in August 1971, and each portfolio (outside of 100% stocks) contains a 5% allocation to gold. A few months ago, I explained how a certain configuration of this asset mix has never lost more than 2.8% per year over the same timeframe considered for this post. The purpose behind gold for most portfolios isn’t because it’s such a great long-term investment. Rather it’s to diversify away from having an excess concentration of your wealth being wrapped up in a single currency. Having all of your personal assets denominated in US dollars or any other currency carries risks similar to having a lot of wealth wrapped up in any other single asset or asset class. Part I discussed how concentration increases risk exponentially, not linearly. What the right allocation of gold might be in a portfolio depends on what other assets are owned, what currencies are owned, gold’s price, and other such matters. Because gold is not a particularly high-yielding investment over the long-run, the right allocation is not particularly high. This is usually less than 10%, or the point where reward-to-risk will be maximized. For more on gold specifically, this post discusses it in more detail. ___ The Testing Setup & Future Expected Asset Class Returns Here are the portfolios that will be used. There are 21 in total: Backtesting will not be used for this exercise. Why? Because future returns will be different than past returns. How do we know this? Because the fundamental drivers of asset classes on a macroeconomic scale – growth and inflation – will put out different readings than they did in the past, so the returns of financial assets will be different as well. ___ Stocks Since 1972, equities have returned 11% in nominal terms. Inflation was about 4% annualized, so equities returned 7% in real terms. We know that inflation will be lower in the future because central banks think much differently about price pressures in an economy than they did 3-4 decades ago. Now, 2% annual inflation is considered a reasonable goal, given some inflation will always be present when trying to get economies toward “full employment.” This is due to the structural frictions associated with imperfect competition (e.g., monopoly, duopoly, oligopoly, monopsony). Accordingly, it is virtually impossible to get an economy to full employment without some level of inflation. US stocks will not return 7% in real terms over the long-run going forward. How do we know this? Because growth over the long-run is simply a mechanical function of productivity growth and growth in the number of workers. In the past, productivity growth was higher due to technological advancement and because of higher growth in the workforce. Going forward, productivity growth will average somewhere around 1.7%-2.0%, and growth in the workforce will become increasingly difficult as the population ages. This was a big source of economic growth from the mid-1960s through the beginning of the century due to increased population growth and female participation in the labor market. But it has since remained stagnant and is actually lower than it was twelve years ago. With 2% growth and 2% inflation, you’ll get some additional contribution from dividends. But that contribution won’t be as high as it was it was in the past. So you can call the forward annual returns of stocks about 6%. You can also make a reasonable case for 7% or 5%. ___ Bonds Bonds have returned 6.9%, or about 2.9% in real terms. The forward returns of short-to-intermediate term USD safe credit will be approximately 3% in nominal terms. How is this determined? Because you look at what’s discounted in the yield curve. At 2% inflation, your real return is 1%. ___ Gold Gold historically has provided a slightly better return than cash. (Note: Cash means the risk-free rate, not a 0%-returning asset unless the risk-free rate is indeed 0%.) The price of gold is closely intertwined with the dynamics of the US dollar. This is because of the dollar’s historical relationship as the reserve currency in the Bretton Woods monetary system from 1944 to 1971. This agreement dictated the financial and commercial relations among the world’s developed countries/main military powers at the time, excluding the Soviet Union. The US controlled nearly 70% of the world’s gold at the time, and each country agreed to tie its currency tightly to gold. Over the long-run, the price of gold will approximate the amount of dollar currency and reserves in circulation divided by the global gold stock. The former tends to grow at a faster clip than gold is mined. This dynamic is a basic matter of the demand for currency/reserves versus the demand for physical gold. This is expected to come in around 2% annualized long-term. ___ The rules 1. Portfolios will be leveraged to mirror the volatility of a 100% stocks allocation. For example, since the US went off the gold standard, equities have had an annualized volatility of about 15.3%. Thus, if a particular stocks/bonds/gold allocation has an annualized volatility of 10%, it will be leveraged 1.53x to match it. This is done at the portfolio level, rather than at the asset level. 2. Cost of leverage financing is assumed to be 1.5%. You can’t get USD financing this cheap currently. However, that figure would be considered quite high if you’re borrowing in CHF, EUR, or JPY, and standard for GBP. Currently, in terms of currencies that go from expensive to cheap to borrow in, they go in the following order: USD > GBP > JPY > EUR > CHF 3. As mentioned in the preceding section, the future returns estimate of stocks will be taken as 6%, future returns estimate of bonds as 3%, and the future return of gold as 2%. 4. Rebalance once annually. 5. No market timing. For example, even when it’s classically the best time to buy stocks – high output gap/unemployment, central bank easing monetary policy, wide credit spreads – your allocation to stocks is the same as it would be when it’s classically the worst time to buy stocks – low output gap/unemployment, central bank tightening monetary policy, tight credit spreads. Of course, in real-life you might approach things differently but that’s not the idea of this exercise. ___ Results - Results: Section I Portfolio allocations will be formatted as Stocks/Bonds/Gold. For example, 0/95/5 means 0% stocks / 95% bonds / 5% gold. As mentioned above, I will not provide detailed historical performance stats because the returns are effectively meaningless as it pertains to the future. Instead, I’ll provide volatility (as measured by standard deviation), leverage, worst year, maximum drawdown, and correlation to US stocks. - 0/95/5 Key stats: As abovementioned, leveraging is done at the portfolio level to match equities. Stocks have turned in an annualized volatility of 15.3% since 1972. Leveraging a portfolio with 5.74% volatility to match that of equities would be leveraged by a factor of 2.67x. In other words, for every $1 in equity, $1.67 would be borrowed. - 5/90/5 Key stats: - 10/85/5 Key stats: Note in the statistics that even as we add equities to the portfolio the risk metrics improve – namely, lower standard deviation (permits higher leverage), better worst years, and better maximum drawdowns. This is the diversification effect because you’re reducing the covariance of the revenue streams. Also note that even a 10% allocation to equities drives our correlation to the stock market up to 35%. - Performance These three portfolios are compared to the relative performance to each other, based on their future returns expectations, and a portfolio of 100% stocks. But first, since people tend to naturally be curious, here is what the result would have been if you were applying this from 1972 to 2018: (click to enlarge) And since time series relationships are often difficult to view graphically when compounding effects are involved, this displays the relationship logarithmically: As we can see, the performance is markedly better. This follows one of the concepts in early parts of this series: A diversified portfolio with moderate leverage offers better reward-to-risk than a highly concentrated portfolio with no leverage. This might seem too good to be true at first glance, but this is how achieving superior risk-adjusted returns with moderate leverage, combined with the effects of compounding over time, lead to much higher returns. This divergence becomes very apparent because the data cover 47 years. Since the 10/85/5 portfolio yielded 18.52% annualized in the way it would have been leveraged using leverage cost assumptions of 1.5%, and stocks returned about 11%, if you take the difference in compounding over 47 years, then yes, the difference will be a factor of about 21. This is not a miscalculation. For example, if you improve your return from 5% to 10% per year, you are not simply doubling your return after 47 years because of the compounding effects that go into it. Instead, you are improving the aggregate return by a factor of approximately 9-to-1. But back to the appropriate focus on future returns, if we take this another 47 years into the future, this is what the expected returns will look like: Stocks (“100/0/0”) outdoes 0/95/5 and 5/90/5 because the latter two portfolios don’t provide enough in expected returns even when leveraged to match the volatility of equities. However, when the risk premium in stocks contracts enough to the point where you would be indifferent between investing more in stocks or bonds, you’re at a point where your asset allocation would shift toward bonds and away from stocks if the risk premium were to compress further. If you play the markets actively, quantifying these spreads can be important to ensure you’re getting the most bang for your buck. Based on this simple model, if you were make the same assumptions I’m using regarding future returns, cost of leverage financing, and future volatility, you would prefer the 10/85/5 portfolio over the 100/0/0, though you’d be close to indifferent between 5/95/5 and 100/0/0. ___ Results: Section II - 15/80/5 Key stats: Note that with just 15% allocation to equities, you are already about 50% correlated to the stock market. This is because stocks are much more volatile than intermediate duration safe credit. So once you start getting up toward 20%+ allocation toward equities/risk assets you will find that they start to dominate the movement of your portfolio. The beta risk at a certain point will begin to outweigh the alpha risk – because investors tend to back into a very equity-centric approach and alpha accordingly is difficult to generate – and the overall beta return is increasingly driven by equities. - 20/75/5 Key stats: Another important feature – notice how the maximum drawdowns are less severe as you add equities into this range. This is once again part of the diversification effect. - 25/70/5 Key stats: -Performance All three portfolios perform better on a risk-adjusted basis than 100/0/0, adding 73bps, 98bps, and 113bps of annualized alpha for 15/80/5, 20/75/5, and 25/70/5, respectively. ___ Results: Section III - 30/65/5 Key stats: While maximum drawdowns compressed as we added stocks to the portfolio due to the excessive concentration in bonds, as we get up to a 30% stocks allocation this metric starts expanding again. This is because we start going beyond the point where risk is minimized. Whether taking this extra risk is worth it depends on the type of return achieved in relation. - 35/60/5 Key stats: We are also now past an +80% correlation to the stock market with equities at just 35% of the allocation, which speaks toward the volatility of stocks relative to bonds. - 40/55/5 Key stats: - Performance At this point, we’re starting to see the extra risk in allocating more into equities not outweigh its potential reward relative to the previous set of portfolios. The 30/65/5 portfolio outperforms the 25/70/5 portfolio and slightly outperforms the 35/60/5 portfolio. By 40/55/5, we start to see a decline, which will naturally get more pronounced as we move into the next section. ___ Results: Section IV - 45/50/5 Key stats: We’re now over 90% correlated to the stock market at just a 45% allocation to stocks. This means that the vast majority of our risk concentration is in equities because they’re much more volatile relative to bonds. Also note the maximum drawdown level and how that’s increased significantly. On top of that, the leverage ratio is now less than 2x because of this extra volatility. - 50/45/5 Key stats: - 55/40/5 Key stats: - Performance The extra alpha from these portfolios amounts to about 80-100 bps annually. Maximum drawdowns are also lower (34%-44% as opposed to 51% for 100/0/0). But the outperformance is notably dipping because the imbalance in the portfolio is picking up. This is causing greater volatility and worse long-run performance, both in an absolute and relative sense with respect to portfolios in preceding sections. ___ Results: Section V - 60/35/5 Key stats: - 65/30/5 Key stats: - 70/25/5 Key stats: This is the typical portfolio of pension funds. They tend to seek out the highest returns possible, but have limitations with respect to how much leverage they can use. This lends itself to a strategy where they invest in stocks to a level of risk they can tolerate, then use the remainder to purchase mostly bonds. Going back to this chart from Part II, when leverage is off-limits, mostly off-limits, or simply undesired, then maximizing your returns will entail having to choose asset classes toward the upper right corner of the chart. (click to enlarge) That mostly means investing in assets with leverage already embedded within them, like stocks or real estate, or riskier/leveraged types of equities, such as venture capital or private equity. Even though 25% of the portfolio is in other assets (mostly bonds), the portfolio is still 98% correlated to the stock market because stocks are more volatile than bonds and weighted in a 2x-3x higher allocation. Over the long-run a portfolio like this without leverage would come with the benefits of lower drawdowns (by about one-third) and bad years for the stock market won’t impact you as badly. But the trade-off is a lower annualized return (by close to 100 bps annualized). An example of this type of strategy on WhoTrades Live can be found here: (Source: WhoTrades Live Portfolio page) This trader puts money into equities, with more weight toward IT/technology, and adds some long-duration US bonds and gold through ETFs. - Performance Annualized alpha over the 100% stocks portfolio is now just 45-70 bps. ___ Results: Section VI - 75/20/5 Key stats: - 80/15/5 Key stats: - 85/10/5 Key stats: - Performance Annualized alpha is now in the range of 20-40bps. ___ Results: Section VI - 90/5/5 Key stats: - 95/0/5 Key stats: - 100/0/0 Key stats: - Performance The 90/5/5 is expected to produce just 10bps of alpha per year. The 95/0/5 portfolio is effectively the same as the 100/0/0. - Conclusion This exercise provides basic insight into how diversification and moderate leveraging can provide superior risk-adjusted returns over time relative to the traditional approach where concentration increases risks exponentially. On top of this, the time and maintenance required in this approach is minimal. Rebalancing can be done just once per year, for example, or whenever there is a big move in the market that disrupts your allocation away from the balance that is most in line with your goals. In the traditional approach to asset management where the portfolio’s performance is heavily tied to the performance of equities, this approach has an entirely different type of risk. Namely, instead of being dependent on a good economy with low inflationary pressure like stocks, instead its risk is mainly about its whether it will underperform cash. Investors should feel comfortable with this risk because capitalist economies can’t function if financial assets underperform holding cash for any elongated duration. Moreover, the leverage employed in these portfolios, even the best ones (20/75/5, 25/70/5, 30/65/5, 35/60/5), is moderate, where the leverage ratio is only around 2.5x ($1.50 borrowed for every $1 invested). In short, investment returns can be enhanced if portfolios have diversification among asset classes and can be calibrated toward one’s targeted risk level or desired level of return. You are free to take these concepts or leave them, as there is more than one way to skin a cat. That’s what WhoTrades Live is fundamentally all about as a crowdsourcing investment ideas and sharing them with a broader community of traders and investors. Each profile shows full transparency over virtually everything you need to know to make a decision as to whether it fits your personal needs and expectations. This includes returns, drawdowns, performance trends, holdings and percentage allocations, a computer- and/or user-generated synopsis of the trading style, popularity based on follower count, and any relevant trading activity.