Quantitative Allocation Journal

Discussion in 'Journals' started by ManagedAccount, Feb 28, 2012.

  1. An Introduction

    Who are we?

    My friend and I after years of suffering at the hands of the muppets (yes, you know who you are) have decided to partner up with the hopes of creating a managed account platform. We both come from wonderful pedigrees (e.g., ivy league schools, from an extremely prestigious financial institution and work as quantitative traders). But you don’t care about that, and you shouldn’t.

    What should you care about?

    We are going to make you money.

    I have heard this before. How are you going to make me money?

    Initially, we have created a robust and simple trading strategy that can be implemented in a managed account for a retail investor. At the same time, we have a multiplicity of intermediate and higher frequency strategies, but for the purpose of this journal, we are just going to focus on the simple strategy.
    With respect to that strategy, we are going to trade ONLY the most liquid exchange traded funds. At the end of each month at the close, we are going to exit our existing month’s position and apply our new allocation. Each month we will post our allocation after the close on the last trading day of each month. The portfolio will be long only, and it will apply small amounts of leverage (we will always be at least 100% invested and never more than 200%). Leverage is applied on a proprietary basis depending on our model.

    What is your strategy?

    The strategy is simple. By utilizing a simple orthogonal universe of ETFs that covers the major asset classes (equities, bonds, commodities, volatility, quasi-equities, etc.), we apply our proprietary tactical allocation model on a walk forward optimization basis to derive our monthly weights. Weights and allocations to different ETFs is non-discretionary and completely based on our mathematical model.

    We also have one simple rule: no cherry picking. One of the problems with Wall Street is that they are great at lying to their clients and even better at lying to themselves.

    Finally, we wanted to reiterate that we are quantitative traders, and there is no technical analysis in our model. We do not use any Fibonacci numbers, Bollinger bands or other technical voodoo mumbo jumbo.

    How much are you running?

    For March, we are going to be running $480,000 USD of our own money.

    What are your expected returns?

    We want to achieve annual returns of 20% per year and volatility of less than 10% for a sharpe ratio of at least 2. Our max drawdown should never be more than 10%. Please also remember that we will always be at least 100% invested.

    How long have you been running your strategy?

    We were beta testing for the last six months and running money since the beginning of the year. We have now formalized our strategy and we would like to share it with the community.

    What are this month’s allocations?

    Stay tuned. We will post it tomorrow.
  2. Positions taken as of tonight:
    • AMJ 15.62%
    • DBC 6.50%
    • EMB 32.77%
    • HYG 4.74%
    • IWM 2.06%
    • LQD 35.74%
    • MUB 7.69%
    • SHY 4.87%
    • SPY 26.41%
    • TLT 9.64%
    • VXX 1.52%
    • VXZ 2.44%
    We will keep this updated every 2-4 weeks going forward.
  3. mickson


    How do we know you are not gaming us?
    How often do you rebalance?
    Do we get to see your actual account so that we can verify these are the only trades?:confused:
  4. Had you looked at AMLP or AGG and if so why did you opt out (liquidity?)? Did you also consider IEF?

    Good luck, looks pretty good imo.
  5. @mickson: First, every relationship begins with trust, whether it is a husband and a wife or your financial advisor and your investment goals. Recently, there has been a lot of media coverage about Greg Smith from Goldman, who quit in spectacular fashion from GS(http://www.nytimes.com/2012/03/14/opinion/why-i-am-leaving-goldman-sachs.html). My name is not Greg Smith, but I feel the same way about my clients. Without my clients, I could not feed my family. We have your best interest in mind, and we will be completely transparent with our strategy.

    With respect to your other questions, we rebalance once a month on the last day of the month. Our quantitative research shows that the last day of the month is the best day to rebalance.

    @Soon2BGreat: We did consider AMLP, when we decided to add AMJ. As you accurately point out, AMJ is more liquid, but we chose AMJ instead because, from a structural perspective, AMJ passes through distributions from the reference MLPs more efficiently than AMLP. AMJ is an exchange-traded note issued by JPMorgan, while AMLP is structured as an ordinary corporation. AMLP is not a registered investment company, and therefore, it does not qualify for pass through tax treatment. As a result, AMLP is not as tax efficient (taxes are paid first at a corporate level) and then distributions are passed to investors net of those taxes. AMJ passes through distributions on a tax-free basis. The trade-off in investing in AMJ is the corporate credit risk, and in our humble opinion, JPMorgan is the best bank in the business and it is a trade-off work taking.

    As for AGG, we looked at AGG and given our other bond ETFs, our quantitative model can account for the return distribution and achieve a more efficient result by allocating to similar ETFs of the same asset class (e.g., TLT, SHY, LQD, EMB, HYG and MUB). While AGG’s weights are fixed, we can synthetically and dynamically adjust our weights with our bond universe. The same is true for IEF.

    Can we answer any other questions?
  6. March was a tough month. Our quantitative strategy was down -1.74% for the period beginning on the close of February 29, 2012 to March 30, 2012. The S&P 500 Total Return Index and the HFRX Global Hedge Fund Index were up +3.3% and down -0.02%, respectively. Based on live trading from November 30, 2011, we currently have a sharpe ratio of 1.12, an annualized return of 5.70% and a realized annual volatility of 5.11%. Please see the attached chart below:


    While the market has humbled us in terms of our absolute returns and our risk-adjusted returns this month, we have faith in our model. It should be noted that current performance is completely in line with our backtests. Approximately 1 in 4 months will show a negative performance, which is usually driven by major turning points within the different asset classes. March was such a major turning point. Every major asset class and other developed market equities sold off. Bonds, Commodities, EM equities and volatility all sold off dramatically, while the developed market equities rallied. For the first quarter ending March 31, 2012, Japanese equities were the best performing major asset class in U.S. dollar terms. Even more strikingly odd was the performance of the S&P 500 Total Return Index during the first quarter of 2012. For the last quarter, the S&P 500 Total Return Index had an annualized return of 45.5% and a sharpe ratio of 4.91. This is unprecedented in the history of S&P.

    For this month, we are rebalancing on the first business day of the month based on our purely quantitative strategy. As of the close of April 2, 2012, our model portfolio has moved into the following allocations:

    • EMB (23.2890%)
    • IWM (26.6112%)
    • LQD (21.9122%)
    • MUB (2.51050%)
    • SHY (21.8685%)
    • SPY (34.09580%)
    • TLT (12.01410%)
    • VNQ (32.51930%)

    We are using leverage equal to approximately 174.2% of capital, and we have closed last month’s positions in AMJ, DBC, HYG, VXX and VXZ.

    As always, questions and comments are very welcome!
  7. interesting, tagging this for later.

    Curious though, why you are leveraged and holding SHY - short term treasury. Surely this gives you a negative arb on your interest rates, and cap appreciation in ST treas would have quite a hurdle to clear each month and be a drag on your results. What's your financing cost if I may ask?

  8. Our quantitative strategy was up 1.1% for the period beginning on the close of March 30, 2012 to the close of April 30, 2012. The S&P 500 Total Return Index and the HFRX Global Hedge Fund Index were up +3.3% and down -0.02%, respectively. Based on live trading from November 30, 2011, we currently have a sharpe ratio of 0.96, with an annualized return of 7.43% and a realized annual volatility of 7.77%. Please see the attached chart below:


    We rebalanced yesterday and as of the close of April 30, 2012, our model portfolio has moved into the following allocations:
    • AMJ (12.16%)
    • EMB (43.74%)
    • LQD (30.54%)
    • SPY (21.88%)
    • TLT (29.83%)
    • VNQ (34.77%)
    • VXX (0.63%)
    We are using leverage equal to approximately 174.03% of capital, and we have closed last month’s positions in IWM and MUB. We scaled our exposure down in SPY and scaled up our exposure to EMB, LQD, TLT and VNQ. We initiated a position in AMJ and VXX.

    With respect to last month’s allocation to SHY:

    At any given point, our model decides whether it should incorporate SHY (as well as other treasury ETFs) given the concurrent financing rate. This month for example SHY returned an annualized 2.3%, well above the 1.4% financing rate that we are facing. If our model decides not to trade an ETF for these reasons, it will still incorporate it in the calculations, but the final exposure will be zero. It is thus treated as 'dummy variable', resulting in the variable leverage (between 100-200%) that we mentioned in one of the earlier posts.